Download Full guideline

Update of clinical guidelines 18 and 34
Hypertension
The clinical management of primary hypertension in
adults
Clinical Guideline 127
Methods, evidence, and recommendations
August 2011
Commissioned by the National Institute for
Health and Clinical Excellence
Hypertension (partial update)
Contents
Published by the National Clinical Guideline Centre at
The Royal College of Physicians, 11 St Andrews Place, Regents Park, London, NW1 4BT
First published 2004, republished 2006
© National Clinical Guideline Centre - 2011
Apart from any fair dealing for the purposes of research or private study, criticism or review, as
permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may
be reproduced, stored or transmitted in any form or by any means, without the prior written
permission of the publisher or, in the case of reprographic reproduction, in accordance with the
terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning
reproduction outside the terms stated here should be sent to the publisher at the UK address
printed on this page.
The use of registered names, trademarks, etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the relevant laws and
regulations and therefore for general use.
The rights of National Clinical Guideline Centre to be identified as Author of this work have
been asserted by them in accordance with the Copyright, Designs and Patents Act, 1988.
Update information
November 2016: A footnote was added to recommendations 46, 47, 53, 55, 57 and 58 covering
angiotensin-converting enzyme (ACE) inhibitors about 2 MHRA drug safety alerts. These alerts
cover ACE inhibitor use during pregnancy and breastfeeding.
Hypertension (partial update)
Contents
Contents
Rationale for update
6
Guideline development group members 7
Acknowledgments 8
Acronyms and abbreviations
9
1
Introduction 10
2
Development of the guideline
12
2.1 What is a NICE clinical guideline? 12
2.2 Who developed this guideline? 13
2.3 What this guideline covers 13
2.4 What this guideline does not cover
13
2.5 Relationships between the guideline and other NICE guidance
2.5.1
Related guidance 14
3
2011 Methods
14
3.1 Developing the review questions and outcomes 14
3.2 Searching for evidence
15
3.2.1
Clinical literature search 15
3.2.2
Health economic literature search 15
3.2.3
Evidence of effectiveness 16
3.2.4
Inclusion/exclusion 16
3.2.5
Methods of combining clinical studies 16
3.2.6
Appraising the quality of evidence by outcomes 17
3.2.7
Grading the quality of clinical evidence 18
3.2.8
Study limitations
18
3.2.9
Inconsistency
19
3.2.10 Indirectness 19
3.2.11 Imprecision 19
3.2.12 Prognostic studies 21
3.3 Evidence of cost-effectiveness 21
3.3.1
Literature review
21
3.3.2
Undertaking new health economic analysis
23
3.3.3
Cost-effectiveness criteria 23
3.4 Developing recommendations
24
3.4.1
Research recommendations
24
3.4.2
Validation process 24
3.4.3
Updating the guideline
24
3.4.4
Disclaimer 24
3.4.5
Funding
25
4
2004 Methods
25
4.1.1
Review methods
25
4.1.2
Group process
27
4.1.3
Evidence statements and recommendations
27
4.1.4
Costs and consequences 29
4.2 2006 methods
30
4.2.1
Clinical evidence
30
4.2.2
Cost-effectiveness evidence
31
5
Guideline summary 32
5.1 Algorithms 32
5.2 Key priorities for implementation 34
5.3 Full list of recommendations
35
1
14
Hypertension (partial update)
Contents
5.4 Key research recommendations 39
Measuring blood pressure 40
6.1 Techniques for measuring blood pressure
40
6.1.1
Manual blood pressure measurement 40
6.2 Cuffs 41
6.3 Conditions and environment
41
6.4 White Coat Hypertension 42
6.5 Blood pressure measurement devices 43
6.5.1
Mercury sphygmomanometer
43
6.5.2
Aneroid sphygmomanometers
43
6.5.3
Automated devices 43
6.6 Ambulatory blood pressure monitors
43
6.7 Home blood pressure monitors 44
6.8 Recommendations 44
6.9 Research recommendation
45
7
Diagnosis of Hypertension 45
7.1 Predicting outcome using clinic, home and ambulatory measurements
45
7.1.1
Clinical evidence 2004
45
7.1.2
Clinical evidence 2011
46
7.1.3
Evidence statements – clinical
47
7.2 Sensitivity and specificity of clinic, home and ambulatory measurements 51
7.2.1
Clinical evidence
51
7.2.2
Evidence statements – clinical
52
7.3 Cost-effectiveness of clinic, home and ambulatory measurements 54
7.3.1
Economic evidence – literature review 54
7.3.2
Economic evidence - original economic analysis
55
7.3.3
Evidence statements – economic 60
7.4 Measurement protocols for diagnosing hypertension 61
7.4.1
Ambulatory blood pressure measurement
61
7.4.2
Home blood pressure measurement
88
7.5 Link from evidence to recommendations 94
7.6 Recommendations 99
8
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension 100
8.1.1
Hypertension and cardiovascular disease
101
8.2 Routine clinical investigations
101
8.2.1
Urine testing for proteinuria
102
8.2.2
Blood electrolyte, urea, creatinine, glucose and total/HDL cholesterol
levels 102
8.3 Cardiovascular Risk Assessment 102
8.4 Secondary Hypertension 102
8.4.1
Renal and renovascular disease 103
8.4.2
Pheochromocytoma 103
8.4.3
Hyperaldosteronism (primary aldosteronism) 104
8.4.4
Cushing's syndrome
104
8.5 Other identifiable causes of hypertension
104
8.5.1
Hypothyroidism
104
8.5.2
Hyperthyroidism
104
8.5.3
Obstructive sleep apnoea 105
8.5.4
Coarctation of aorta 105
8.5.5
Acromegaly 105
6
2
Hypertension (partial update)
Contents
8.5.6
Drugs 105
8.6 Recommendations 105
8.7 Research recommendations
106
9
Initiating and monitoring treatment, including blood pressure targets 106
9.1 Blood pressure thresholds for initiating pharmacological treatment 106
9.1.1
Clinical evidence
106
9.1.2
Evidence statements - clinical
127
9.1.3
Evidence statements – economic 127
9.2 Treatment of people aged 80 years and greater 127
9.2.1
Clinical evidence
127
9.2.2
Economic evidence 131
9.2.3
Evidence statements – Clinical 131
9.2.4
Evidence statements – Health economic 132
9.3 Link from evidence to recommendations 132
9.4 Recommendations 134
9.5 Recommendations for research 135
9.6 Monitoring treatment efficacy
135
9.6.1
Clinical evidence
135
9.6.2
Economic evidence 141
9.6.3
Evidence statements – clinical
141
9.6.4
Evidence statements – health economic 143
9.6.5
Link from evidence to recommendations 143
9.6.6
Recommendations 144
9.6.7
Research recommendations
144
9.7 Blood pressure targets for treatment
144
9.7.1
Clinical evidence
144
9.7.2
Health economic evidence 158
9.7.3
Evidence statements – clinical
158
9.7.4
Evidence statements – economic 159
9.7.5
Link from evidence to recommendations: blood pressure treatment
targets. 159
9.8 Recommendations 161
9.9 Research Recommendation
162
9.10 Frequency of review
162
10
Integrating the assessment of blood pressure, target organ damage and
cardiovascular risk assessment and clinical decision making regarding treatment
initiation, treatment and targets 162
11
Lifestyle interventions
163
11.1 Overview
163
11.1.1 Managing changes in lifestyle
166
11.1.2 Diet 166
11.1.3 Exercise
168
11.1.4 Relaxation therapies
169
11.1.5 Multiple lifestyle interventions
171
11.1.6 Alcohol
173
11.1.7 Coffee 174
11.1.8 Reducing sodium (salt) intake
174
11.1.9 Calcium supplements
175
11.1.10 Magnesium supplements 176
11.1.11 Potassium supplementation
177
11.1.12 Combined salt supplements
178
3
Hypertension (partial update)
Contents
11.1.13 Drug therapy versus lifestyle change
179
11.1.14 Smoking cessation 180
11.1.15 Recommendations 181
12
Pharmacological interventions
181
12.1 2004 guidance: pharmacological interventions 184
12.1.1 Placebo controlled trials 184
12.2 2006 rapid pharmacological update: head to head trials
191
12.2.1 Clinical evidence statements: head-to-head drug comparisons
191
12.2.2 Meta-analysis results summary 194
12.3 2011 update: Pharmacological therapy for hypertension
196
12.3.1 Angiotensin-converting enzyme inhibitors (ACEi) versus Angiotensin
Receptor Blockers (ARB)
196
12.3.2 Diuretics
200
12.4 Cost-effectiveness analysis
225
12.4.1 Methodological introduction
225
12.4.2 Results of the health economic model 227
12.4.3 Conclusions 228
12.5 Step two therapy
229
12.6 Resistant hypertension
232
12.7 Special groups for consideration 235
12.7.1 People aged over 80 years 235
12.7.2 Younger people
235
12.7.3 Ethnicity
235
12.7.4 Chronic kidney disease
241
12.7.5 Type 1 and Type 2 diabetes
241
12.7.6 Women who are pregnant or breast-feeding
242
12.8 Stopping treatment 242
12.9 Link from evidence to recommendations- Pharmacological treatment of
hypertension
244
12.10
Recommendations 252
12.11
Research recommendations
254
13
Patients’ perspectives
254
13.1 Introduction 254
13.2 Discovering hypertension 254
13.3 Treatment 255
13.4 Living with hypertension 255
13.5 Education and adherence 256
13.5.1 Compliance with Prescribed Antihypertensive Medication
256
13.5.2 Implementing lifestyle measures 257
13.5.3 Recommendations 258
14
Reference list
258
15
Glossary
300
Appendices 309
Appendix A:
Appendix B:
Appendix C:
Appendix D:
Appendix E:
Appendix F:
Appendix G:
Scope 309
Declarations of Interest
317
Review questions 320
Literature search strategies
321
Review protocols 345
Clinical evidence tables
360
Evidence tables – health economic studies (2011 update)
4
368
Hypertension (partial update)
Contents
Appendix H:
Forest plots 373
Appendix I: Cost-effectiveness analysis – pharmacological treatment (updated
2011)
401
Appendix J: Cost-effectiveness analysis – blood pressure monitoring for confirming
a diagnosis of hypertension (new 2011) 427
Appendix K:
Research recommendations (2011)
484
Appendix L:
CG18 Essential Hypertension: managing adult patients in
primary care, 2004 486
Appendix M: CG34 Hypertension: management in adults in primary care:
pharmacological update, 2006 487
5
Hypertension (partial update)
Introduction
Rationale for update
6
Update 2011
This document is a partial update of Clinical Guideline 18 (2004) and Clinical Guideline 34
(2006) on Essential Hypertension in adults. The sections that have not been amended are
integrated with the updated guidance in this document. Both guidelines are available in full in
the appendices of the document.
Improvements in methodology since 2006 mean the way information is presented may, at
times, be inconsistent (for example, the style of review write-up, and 2011 recommendations
are not graded according to the strength of evidence, unlike those in the 2006).
New or amended sections of the guideline are indicated with an ‘update’ panel in the right
hand margin.
Hypertension (partial update)
Introduction
Guideline development group members
Role
Bernard Higgins
Clinical Director, National Clinical Guideline Centre
Bryan Williams
Professor of Medicine, Guideline Development Group Chair
Helen Williams
Consultant Pharmacist for cardiovascular disease,
Southwark Health and Social Care
Jane Northedge
Patient and care representative
John Crimmins
General Practitioner, Vale of Glamorgan
Kate Lovibond
Senior Health Economist, National Clinical Guideline Centre
Mark Caulfield
Professor of Clinical Pharmacology, Barts and the London School of
Medicine
Michaela Watts
Hypertension Nurse Specialist, Addenbrooke’s Hospital, Cambridge
Naomi Stetson
Primary Care Nurse, Watling Medical Centre, Burnt Oak
Paul Miller
Senior Information Scientist, National Clinical Guideline Centre
Rachel O’Mahony
Senior Research Fellow, National Clinical Guideline Centre
Richard McManus
Professor of Primary Care Cardiovascular Research, University of
Birmingham
Shelley Mason
Patient and carer representative
Terry McCormack
General Practitioner, Spring Vale Medical Centre, North Yorkshire, Primary
Care Cardiovascular Society
Taryn Krause
Senior Project Manager/Research Fellow, National Clinical Guideline Centre
7
Update 2011
Name
Hypertension (partial update)
Introduction
Acknowledgments
8
Update 2011
The development of this guideline was greatly assisted by the following people:
 Jill Cobb, Information Scientist, National Clinical Guideline Centre
 Ralph Hughes, Health Economist, National Clinical Guideline Centre
 Fatema Limbada, Project Coordinator, National Clinical Guideline Centre
 Jill Parnham, Director of Operations, National Clinical Guideline Centre
 David Wondering, Health Economic Lead, National Clinical Guideline Centre
 Jacoby Patterson, Systematic Reviewer
 Julie Brown, Systematic Reviewer
 Sue Jowett, Senior Lecturer in Health Economics, University of Birmingham
 James Hodgkinson, Research Fellow, University of Birmingham
 Jonathan Mant, Professor of Primary Care Research, University of Cambridge
 Una Martin, Reader in Clinical Pharmacology, University of Birmingham
 Carl Heneghan, Reader in Evidence-Based Medicine, University of Oxford
 Richard Hobbs, Head of Primary Care Clinical Sciences, University of Birmingham.
Hypertension (partial update)
Introduction
Acronyms and abbreviations
ABPM
ACEi
ANOVA
ARB
BNF
CBPM
CCA
CCB
CEA
c.f.
CI / 95% CI
CUA
DH
DSA
ED
EQ-5D
GDG
GP
GRADE
Evaluation
HBPM
HES
HR
HRQoL
HT
HTA
ICD-10
ICER
ICH
ISH
IQR
Ambulatory blood pressure measurement
Angiotensin-converting enzyme inhibitors
Analysis of variance
Angiotensin receptor blocker
British National Formulary
Clinic blood pressure measurement
Cost-consequences analysis
Calcium channel blocker
Cost-effectiveness analysis
Confer (refer to)
Confidence interval / 95% confidence interval
Cost-utility analysis
Department of Health
Deterministic Sensitivity Analysis
Emergency Department
EuroQol-5D
Guideline Development Group
General Practitioner
Grading of Recommendations Assessment, Development and
INMB
Incremental Net Monetary Benefit
IRR
Inter-rater reliability
ITT
Intention to treat
LOS
Length of Stay
LR+
Positive likelihood ratio
LY
Life-year
MD
Mean difference
NCGC
National Clinical Guideline Centre
NHS
National Health Service
NHSEED
The NHS Economic Evaluation Database
NICE
National Institute for Health and Clinical Excellence
Home blood pressure measurement
Hospital Episode Statistics
Hazard Ratio
Health-related quality of life
Hypertensive / hypertension
Health technology assessment
International Classification of Diseases, 10th edition
Incremental cost-effectiveness ratio
Isolated clinic hypertension
Ischemia
Interquartile range
9
Hypertension (partial update)
Introduction
1
NNT
Number needed to treat
NPV
Negative predictive value
NS
Non-significant (not statistically significant)
NT
Normotensive
OR
Odds ratio
PICO
Framework incorporating patients, interventions, comparison and outcome
PPP
Purchasing Power Parity
PPV
Positive predictive value
p.r.n
Pro re nata
PSA
Probabilistic sensitivity analysis
QALY
Quality-adjusted life year
QUADAS
Quality assessment tool for diagnostic accuracy studies
RCT
Randomised controlled trial
ROC
Receiver operating characteristic
RRK
Riva-Rocci Korotkoff
RR
Relative risk
SD
Standard deviation
SE
Standard error
SPC
Summary of product characteristics
SR
Systematic review
SS
WCH
Statistically significant
White coat hypertension
Introduction
This guideline is for the clinical management of primary hypertension in adults (aged greater
than 18 years). Hypertension (high blood pressure) is one of the most preventable causes of
premature morbidity and mortality world-wide.
Hypertension is a major risk factor for stroke (ischaemic and haemorrhagic), myocardial
infarction, heart failure, chronic kidney disease, peripheral vasculardisease, cognitive decline
and premature death. Untreated hypertension is associated a progressive rise in blood
pressure, often culminating in a treatment resistant state due to associated vascular and renal
damage.
Blood pressure is quantified as diastolic and systolic pressures measured in millimetres of
mercury (mmHg). The diastolic pressure represents the pressure during ventricular relaxation
in diastole whereas the systolic pressure represents the peak pressure due to ventricular
contraction during systole. Either or both pressures have specified upper limits of normal and
elevation in either or both pressures are used to define hypertension.
Blood pressure is normally distributed in the population and there is no natural cut-point
above which "hypertension" definitively exists and below which, it does not. Epidemiological
studies demonstrate that the aforementioned disease risk associated with blood pressure is a
10
Hypertension (partial update)
Introduction
11
Update 2011
continuous relationship and above blood pressures of 115/70mmHg, the risk of cardiovascular
events doubles for every 20/10mmHg rise in blood pressure. The threshold blood pressure
determining the presence of hypertension is defined as the level of blood pressure above
which treatment has been shown to reduce the development or progression of disease.
Primary hypertension was previously termed “essential hypertension” because of a longstanding view that high blood pressure was sometimes “essential” to perfuse diseased and
sclerotic arteries. It is now recognised that the diseased and sclerotic arteries were most often
the consequence of the hypertension and thus the term “essential hypertension” is redundant
and the “primary hypertension” is preferred. Primary hypertension refers to the majority of
people with sustained high blood pressure (approximately 90%) encountered in clinical
practice, for which there is no obvious, identifiable cause. The remaining 10% are termed
"secondary hypertension" for which specific causes for the blood pressure elevation can be
determined (for example, Conn's adenoma, renovascular disease, or phaeochromocytoma).
Primary hypertension is remarkably common in the UK population and the prevalence is
strongly influenced by age and lifestyle factors. Systolic and/or diastolic blood pressures may
be elevated. Systolic pressure elevation is the more dominant feature of hypertension in older
patients and diastolic pressure more commonly elevated in younger patients, (those less than
50 years of age). At least one quarter of the adult population of the UK have hypertension,
(blood pressure ≥140/90mmHg) and more than half of those over the age of 60 years. As the
demographics of the UK shifts towards an older, more sedentary and obese population, the
prevalence of hypertension and its requirement for treatment will continue to rise.
Routine periodic screening for high blood pressure is now commonplace in the UK as part of
National Service Frameworks for cardiovascular disease prevention. Consequently, the
diagnosis, treatment and follow-up of patients with hypertension is one of the most common
interventions in primary care, accounting for approximately 12% of Primary Care
consultation episodes and approximately £1 billion in drug costs in 2006 .
NICE first issued guidance for the management of hypertension in primary care in 2004. This
was followed by a rapid update of the pharmacological treatment chapter of the guideline in
2006. The current partial update of the hypertension guideline is in response to the regular
five year review cycle of existing NICE guidance. It began with a scoping exercise which
identified key areas of the existing guideline for which new evidence had emerged that was
likely to influence or change existing guideline recommendations.
Sections of the guideline that have not been updated continue to stand, however, wherever
NICE has subsequently issued new and related guidance relevant to existing
recommendations, these have been identified and cross-referred to in this partial update,
examples include interventions on lifestyle factors and public health policy recommendations
such as smoking cessation, dietary salt restriction, alcohol intake and cardiovascular disease
prevention and cardiovascular disease risk assessment. In addition, new NICE guidance
developed in areas relevant to hypertension are also highlighted and cross referenced (for
example, chronic kidney disease, stroke, diabetes and hypertension in pregnancy).
The recommendations that have been reviewed in this partial update of the guideline for the
clinical management of primary hypertension in adults, include; blood pressure measurement
for the diagnosis of hypertension; blood pressure thresholds for intervention with drug therapy
and blood pressure targets for treatment; specific aspects of the recommendations for the
pharmacological treatment of hypertension; the treatment of hypertension in the very elderly
(people aged greater than 80 years); dilemmas surrounding decision making for treatment of
hypertension in younger adults (less than 40 years); the treatment of drug resistant
hypertension; and wherever appropriate, the impact of age and ethnicity on treatment
recommendations.
Finally, despite the fact that the treatment of hypertension has a large clinical trial evidence
base to inform recommendations, an important aspect of the evidence review for guideline
Hypertension (partial update)
Development of the guideline
development is to identify where gaps in knowledge remain. In so doing, research questions
have been identified to prompt the gathering of further evidence to continue the evolution of
guidance and clinical practice.
2
Development of the guideline
2.1 What is a NICE clinical guideline?
12
Update 2011
NICE clinical guidelines are recommendations for the care of individuals in specific clinical
conditions or circumstances within the NHS – from prevention and self-care through primary
and secondary care to more specialised services. We base our clinical guidelines on the best
available research evidence, with the aim of improving the quality of health care. We use
predetermined and systematic methods to identify and evaluate the evidence relating to
specific review questions.
NICE clinical guidelines can:
 provide recommendations for the treatment and care of people by health professionals
 be used to develop standards to assess the clinical practice of individual health
professionals
 be used in the education and training of health professionals
 help patients to make informed decisions
 improve communication between patient and health professional
While guidelines assist the practice of healthcare professionals, they do not replace their
knowledge and skills.
We produce our guidelines using the following steps:
 Guideline topic is referred to NICE from the Department of Health
 Stakeholders register an interest in the guideline and are consulted throughout the
development process.
 The scope is prepared by the National Clinical Guideline Centre (NCGC)
 The NCGC establishes a guideline development group
 A draft guideline is produced after the group assesses the available evidence and makes
recommendations
 There is a consultation on the draft guideline.
 The final guideline is produced.
The NCGC and NICE produce a number of versions of this guideline:
 The full guideline contains all the recommendations, plus details of the methods used and
the underpinning evidence
 The NICE guideline lists the recommendations
 the Quick Reference Guide (QRG) presents recommendations in a suitable format for
health professionals
 Information for the public - ‘understanding NICE guidance’ or UNG - is written using
suitable language for people without specialist medical knowledge
 Clinical Pathway – www.pathways.nice.org.uk/pathways/hypertension
This version is the full guideline. The other documents can be downloaded from NICE at
www.nice.org.uk
Hypertension (partial update)
Development of the guideline
2.2 Who developed this guideline?
A multidisciplinary Guideline Development Group (GDG) comprising professional group
members and consumer representatives of the main stakeholders developed this guideline (see
section on Guideline Development Group Membership and acknowledgements).
The National Institute for Health and Clinical Excellence funds the National Clinical
Guideline Centre (NCGC) and thus supported the development of this guideline. The GDG
was convened by the NCGC and chaired by Professor Bryan Williams in accordance with
guidance from the National Institute for Health and Clinical Excellence (NICE).
The group met every four weeks during the development of the guideline. At the start of the
guideline development process all GDG members declared interests including consultancies,
fee-paid work, share-holdings, fellowships and support from the healthcare industry. At all
subsequent GDG meetings, members declared arising conflicts of interest, which were also
recorded in Appendix B: Declarations of Interest.
Members were either required to withdraw completely or for part of the discussion if their
declared interest made it appropriate. The details of declared interests and the actions taken
are shown in Appendix B: Declarations of Interest.
Staff from the NCGC provided methodological support and guidance for the development
process. The team working on the guideline included a project manager, systematic
reviewers, health economists and information scientists. They undertook systematic searches
of the literature, appraised the evidence, conducted meta analysis and cost effectiveness
analysis where appropriate and drafted the guideline in collaboration with the GDG.
2.3 What this guideline covers
2.4 What this guideline does not cover




People with diabetes.
Children and young people (younger than 18 years).
Pregnant women.
Secondary causes of hypertension (for example, Conn's adenoma, phaeochromocytoma
and renovascular hypertension).
13
Update 2011
 Adults with hypertension (18 years and older).
 Particular consideration will be given to the needs of black people of African and
Caribbean descent and minority ethnic groups where these differ from the needs of the
general population.
 People aged 80 years or older.
 Ambulatory monitoring.
 Home blood pressure monitoring.
 Blood pressure thresholds for intervention and targets for treatment.
 First-line therapy options, for example angiotensin-converting enzyme inhibitors versus
angiotension receptors blockers.
 Calcium-channel blockers versus diuretics as preferred components in step two of the
treatment algorithm, for example, combination therapy.
 Adherence to medication.
 Provision of appropriate information and support.
 Resistant hypertension (that is, fourth-line therapy).
 Response to blood pressure lowering drugs according to age and ethnicity.
For further details please refer to Appendix A: Scope and Appendix C: Review questions.
Hypertension (partial update)
2011 Methods
 People with accelerated hypertension (that is, severe acute hypertension associated grade
III retinopathy and encephalopathy).
 People with acute hypertension or high blood pressure in emergency care settings.
 Prevention of hypertension.
 Screening for hypertension.
 Specialist management of secondary hypertension (that is, hypertension arising from other
medical conditions).
 Non-pharmacological interventions.
2.5 Relationships between the guideline and other NICE guidance
2.5.1
2011 Methods
This guidance was developed in accordance with the methods outlined in the NICE
Guidelines Manual 2009.430
3.1 Developing the review questions and outcomes
Review questions were developed in a PICO framework (patient, intervention, comparison
and outcome) for intervention reviews, and with a framework of population, index tests,
reference standard and target condition for reviews of diagnostic test accuracy. This was to
guide the literature searching process and to facilitate the development of recommendations
by the guideline development group (GDG). They were drafted by the NCGC technical team
14
Update
2011
3
Related guidance
 Prevention of cardiovascular disease at the population level. NICE Public Health Guidance
25/ www.nice.org.uk/PH25
 Medicines adherence. NICE clinical guideline 76 (2009). Available from
www.nice.org.uk/guidance/CG76
 Chronic kidney disease. NICE clinical guideline 73 (2008). Available from
www.nice.org.uk/guidance/CG73
 Stroke. NICE clinical guideline 68 (2008). Available from
www.nice.org.uk/guidance/CG68
 Lipid modification. NICE clinical guideline 67 (2008). Available from
www.nice.org.uk/guidance/CG67
 Type II diabetes. NICE clinical guideline 66 (2008). Available from
www.nice.org.uk/guidance/CG66
 Sleep apnoea – continuous positive airway pressure (CPAP). NICE technology appraisal
guidance 139 (2008). Available from www.nice.org.uk/guidance/TA139
 MI: secondary prevention. NICE clinical guideline 48 (2007). Available from
www.nice.org.uk/guidance/CG48
 Obesity. NICE clinical guideline 43 (2006). Available from
www.nice.org.uk/guidance/CG43
 Atrial fibrillation. NICE clinical guideline 36 (2006). Available from
www.nice.org.uk/CG36
 Nutrition support in adults. NICE clinical guideline 32 (2006). Available from
www.nice.org.uk/guidance/CG32
 Chronic heart failure. NICE clinical guideline 5 (2003). Available from
www.nice.org.uk/guidance/CG5
Hypertension (partial update)
2011 Methods
and refined and validated by the GDG. The questions were based on the key clinical areas
identified in the scope (Appendix A: Scope) and a list can be found in Appendix C: Review
Questions. Further information on the outcome measures examined follows this section.
3.2 Searching for evidence
3.2.1
3.2.1.1
3.2.2
Clinical literature search
Systematic literature searches were undertaken to identify evidence within published literature
in order to answer the review questions as per The Guidelines Manual (2009).430 Clinical
databases were searched using relevant medical subject headings, free-text terms and study
type filters where appropriate. Studies published in languages other than English were not
reviewed. All searches were conducted on core databases, MEDLINE, Embase, Cinahl and
The Cochrane Library. All searches were updated on 29th November 2010 and no papers
were included beyond this date.
Search strategies were checked by looking at reference lists of relevant key papers, checking
search strategies in other systematic reviews and asking the GDG for known studies. The
questions, the study types applied, the databases searched and the years covered can be found
in Appendix C: Literature search strategies.
During the scoping stage, a search was conducted for guidelines and reports on the websites
listed below and via organisations relevant to the topic. Searching for grey literature or
unpublished literature was not undertaken. All references sent by stakeholders were
considered.
 Guidelines International Network database (www.g-i-n.net)
 National Guideline Clearing House (www.guideline.gov/)
 National Institute for Health and Clinical Excellence (NICE) (www.nice.org.uk)
 National Institutes of Health Consensus Development Program (consensus.nih.gov/)
 National Library for Health (www.library.nhs.uk/)
Call for evidence
The GDG decided to initiate a ‘call for evidence’ for meta-analyses, based on a systematic
review, that include studies that use ambulatory blood pressure measurement as the reference
standard and report sensitivity and specificity of home and/or clinic blood pressure
measurement, as they believed that important evidence existed that would not be identified by
the standard searches. The NCGC contacted all registered stakeholders and asked them to
submit any relevant published or unpublished evidence.
Health economic literature search
Systematic literature searches were also undertaken to identify health economic evidence
within published literature relevant to the review questions. The evidence was identified by
conducting a broad search relating to the guideline population in the NHS economic
evaluation database (NHS EED), the Health Economic Evaluations Database (HEED) and
health technology assessment (HTA) databases from 2003 onwards to find anything published
since the original guideline. There were two questions not covered in either the original
guideline or the previous rapid update, for which additional searches with no date restrictions
were carried out. Additionally, the search was run on MEDLINE and Embase, with a specific
economic filter, from 2009, to ensure recent publications that had not yet been indexed by
these databases were identified. Studies published in languages other than English were not
reviewed. Where possible, searches were restricted to articles published in English
language.The search strategies for health economics are included in Appendix D: Literature
search strategies. All searches were updated on 29th November 2010. No papers published
after this date were considered.
15
Updat
Hypertension (partial update)
2011 Methods
3.2.2.1
3.2.3
3.2.4
3.2.5
Call for evidence
The GDG decided to initiate a ‘call for evidence’ for cost-effectiveness analyses from a UK
perspective, using methods in line with the NICE reference case, comparing ambulatory,
home and clinic blood pressure measurement in the diagnosis of hypertension, as they
believed that important evidence existed that would not be identified by the standard searches.
The NCGC contacted all registered stakeholders and asked them to submit any relevant
published or unpublished evidence.
Evidence of effectiveness
The Research Fellow:
 Identified potentially relevant studies for each review question from the relevant search
results by reviewing titles and abstracts – full papers were then obtained.
 Reviewed full papers against pre-specified inclusion / exclusion criteria to identify studies
that addressed the review question in the appropriate population and reported on outcomes
of interest (review protocols are included in Appendix E:Review protocols).
 Critically appraised relevant studies using the appropriate checklist as specified in The
Guidelines Manual 430
 Extracted key information about the study’s methods and results into evidence tables
(evidence tables are included in Appendix D: Evidence tables – clinical studies and
Appendix G: Evidence tables – health economic studies.
 Generated summaries of the evidence by outcome (included in the relevant chapter writeups):
o Randomised studies: meta analysed, where appropriate and reported in GRADE
profiles (for clinical studies) – see below for details
o Observational studies: data has been presented for individual studies narratively or in
summary tables (GRADE profiles have not been generated)
o Diagnostic studies: data has been presented for individual studies narratively or in
summary tables (GRADE profiles have not been generated)
o Qualitative studies: each study summarised in a table where possible, otherwise
presented in a narrative.
Inclusion/exclusion
See the review protocols in Appendix E: Review Protocols for full details.
Methods of combining clinical studies
16
Update 2011
Data synthesis for intervention reviews
Where possible, meta-analyses were conducted to combine the results of studies for each
review question using Cochrane Review Manager (RevMan5) software. Fixed-effects (Mantel
-Haenszel) techniques were used to calculate risk ratios (relative risk) for the following binary
outcomes: angioedema. Where reported, time-to-event data was presented as a hazard ratio
for the following binary outcomes: mortality, stroke, MI, heart failure, new onset diabetes,
vascular procedures, angina requiring hospitalisation, study drug withdrawal. The continuous
outcome blood pressure (mmHg)] was analysed using an inverse variance method for pooling
weighted mean differences and where the studies had different scales, standardised mean
differences were used. No quality of life outcome data was reported by any of the studies
included in the 2012 update reviews
Statistical heterogeneity was assessed by considering the chi-squared test for significance at
p<0.1 or an I-squared inconsistency statistic of >50% to indicate significant heterogeneity.
Where significant heterogeneity was present, we carried out sensitivity analysis based on the
quality of studies, with particular attention paid to allocation concealment, blinding and loss
Hypertension (partial update)
2011 Methods
3.2.6
Table 1: Description of quality elements in GRADE for intervention studies.
Quality element
Description
Limitations
Limitations in the study design and implementation may bias the estimates of the
treatment effect. Major limitations in studies decrease the confidence in the estimate of
the effect.
Inconsistency
Inconsistency refers to an unexplained heterogeneity of results.
Indirectness
Indirectness refers to differences in study population, intervention, comparator and
outcomes between the available evidence and the review question, or recommendation
made.
Imprecision
Results are imprecise when studies include relatively few patients and few events and
17
Update 2011
to follow-up (missing data). In cases where there was inadequate allocation concealment,
unclear blinding, high loss to follow-up (≥ 20% missing data for studies ≤2 years follow-up
and ≥30% for those with >2 years follow-up) or differential missing data, this was examined
in a sensitivity analysis. For the latter, the duration of follow up was also taken into
consideration prior to including in a sensitivity analysis.
Assessments of potential differences in effect between subgroups were based on the chisquared tests for heterogeneity statistics between subgroups. If no sensitivity analysis was
found to completely resolve statistical heterogeneity then a random effects (DerSimonian and
Laird) model was also explored to provide a more conservative estimate of the effect.
The means and standard deviations of continuous outcomes were required for meta-analysis.
However, in cases where standard deviations were not reported, the standard error was
calculated if the p-values or 95% confidence intervals were reported and meta-analysis was
undertaken with the mean and standard error using the generic inverse variance method in
Cochrane Review Manager (RevMan5) software. Where p values were reported as “less
than”, a conservative approach was undertaken. For example, if the p value was reported as “p
≤0.001”, the calculations for standard deviations will be based on a p value of 0.001. If these
statistical measures were un available then the methods described in section 16.1.3 of the
Cochrane Handbook ‘Missing standard deviations’ were applied as the last resort.
Appraising the quality of evidence by outcomes
The evidence for outcomes from the included RCT studies were evaluated and presented
using an adaptation of the ‘Grading of Recommendations Assessment, Development and
Evaluation (GRADE) toolbox’ developed by the international GRADE working group
(http://www.gradeworkinggroup.org/). The software (GRADEpro) developed by the GRADE
working group was used to assess the quality of each outcome, taking into account individual
study quality and the meta-analysis results. The summary of findings was presentedas an
‘evidence profile,’ a single table that includes details of the quality assessment as well as
pooled outcome data, where appropriate, an absolute measure of intervention effect and the
summary of quality of evidence for that outcome. In this table, the columns for intervention
and control indicate the sum of the sample size for continuous outcomes. For binary outcomes
such as number of patients with an adverse event, the event rates (n/N: number of patients
with events divided by sum of number of patients) are shown with percentages. Reporting or
publication bias was only taken into consideration in the quality assessment and included in
the Clinical Study Characteristics table if it was apparent.
Each outcome was examined separately for the quality elements listed and defined in Table 1
and each graded using the quality levels listed in Table 2: The main criteria considered in the
rating of these elements are discussed below (see 3.2.7 Grading of Evidence). Footnotes were
used to describe reasons for grading a quality element as having serious or very serious
problems. The ratings for each component were summed to obtain an overall assessment for
each outcome.
GRADE is currently designed only for randomised trials and observational studies.
Hypertension (partial update)
2011 Methods
Quality element
Description
thus have wide confidence intervals around the estimate of the effect relative to the
clinically important threshold.
Publication bias
Publication bias is a systematic underestimate or an overestimate of the underlying
beneficial or harmful effect due to the selective publication of studies.
Table 2: Levels of quality elements in GRADE
Level
Description
None
There are no serious issues with the evidence
Serious
The issues are serious enough to downgrade the outcome evidence by one level
Very serious
The issues are serious enough to downgrade the outcome evidence by two levels
Table 3: Overall quality of outcome evidence in GRADE
3.2.7
Level
Description
High
Further research is very unlikely to change our confidence in the estimate of effect
Moderate
Further research is likely to have an important impact on our confidence in the estimate
of effect and may change the estimate
Low
Further research is very likely to have an important impact on our confidence in the
estimate of effect and is likely to change the estimate
Very low
Any estimate of effect is very uncertain
Table 4: Study limitations of randomised controlled trials
Limitation
Explanation
Allocation
concealment
Those enrolling patients are aware of the group to which the next enrolled patient will
be allocated (major problem in “pseudo” or “quasi” randomised trials with allocation
by day of week, birth date, chart number, etc)
Lack of blinding
Patient, caregivers, those recording outcomes, those adjudicating outcomes, or data
18
Update 2011
Grading the quality of clinical evidence
After results were pooled, the overall quality of evidence for each outcome was considered.
The following procedure was adopted when using GRADE:
1. A quality rating was assigned, based on the study design. RCTs start HIGH and
observational studies as LOW.
2. The rating for RCTs was then downgraded for the specified criteria: Study limitations,
inconsistency, indirectness, imprecision and reporting bias. These criteria are detailed
below. Due to the wide diversity of study design, data reported and data analysis methods
of the observational studies that were included in this guideline, it was very difficult to
compare studies for quality and therefore observational studies were not downgraded or
upgraded in GRADE, and all remained as LOW quality evidence (please see below,
section 0, for details of quality assessment of prognostic studies)..
3. The downgraded marks were then summed and the overall quality rating was revised. For
example, all RCTs started as HIGH and the overall quality became MODERATE, LOW or
VERY LOW if 1, 2 or 3 points were deducted respectively.
4. The reasons or criteria used for downgrading were specified in the footnotes.
The details of criteria used for each of the main quality element are discussed further in the
following sections 3.3.5 to 3.3.8/3.3.9 [if section for publication bias is relevant].
Study limitations
The main limitations for randomised controlled trials are listed in Table 4.
Hypertension (partial update)
2011 Methods
3.2.8
3.2.10
Explanation
analysts are aware of the arm to which patients are allocated
Incomplete
accounting of
patients and
outcome events
Loss to follow-up not accounted and failure to adhere to the intention to treat principle
when indicated
Selective outcome
reporting
Reporting of some outcomes and not others on the basis of the results
Other limitations
For example:
 Stopping early for benefit observed in randomised trials, in particular in the absence
of adequate stopping rules
 Use of unvalidated patient-reported outcomes
 Carry-over effects in cross-over trials
 Recruitment bias in cluster randomised trials
Inconsistency
Inconsistency refers to an unexplained heterogeneity of results. When estimates of the
treatment effect across studies differ widely (i.e. heterogeneity or variability in results), this
suggests true differences in underlying treatment effect. When heterogeneity exists (Chi
square p<0.1 or I- squared inconsistency statistic of >50%), but no plausible explanation can
be found, the quality of evidence was downgraded by one or two levels, depending on the
extent of uncertainty to the results contributed by the inconsistency in the results.
If inconsistency could be explained based on pre-specified subgroup analysis, the GDG took
this into account and considered whether to make separate recommendations based on the
identified explanatory factors, i.e. population and intervention. Where subgroup analysis gave
a plausible explanation of heterogeneity, the quality of evidence was not downgraded.
Indirectness
Directness refers to the extent to which the populations, intervention, comparisons and
outcome measures are similar to those defined in the inclusion criteria for the reviews.
Indirectness is important when these differences are expected to contribute to a difference in
effect size, or may affect the balance of harms and benefits considered for an intervention.
Imprecision
The criteria applied for imprecision are based on the confidence intervals for pooled or the
best estimate of effect as illustrated in Figure 1 and outlined in Table 5.
Update 2011
3.2.9
Limitation
Table 5: Criteria applied to determine precision
Dichotomous and continuous outcomes
The 95% confidence interval (or alternative estimate of precision) around the pooled or best estimate of effect:
Does not cross either of the two minimal important difference (MID) thresholds (the threshold lines for
appreciable benefit or harm); defined as precise
Rating for precision: ‘no serious imprecision’
2.
Crosses one of the two MID thresholds (appreciable benefit or appreciable harm); defined as imprecise
Rating for precision: ‘serious’
3.
Crosses both of the two MID thresholds ( appreciable benefit and appreciable harm); defined as
imprecise
Rating for precision: ‘very serious’
19
Update 2011
1.
Hypertension (partial update)
2011 Methods
Figure 1: Illustration of precise and imprecise outcomes based on the confidence
interval of outcomes in a forest plot
MID
MID
NO SERIOUS
IMPRECISION
SERIOUS
IMPRECISION
-1
VERY
SERIOUS
IMPRECISION
-2
1.0
1.25
Appreciable benefit Non-appreciable benefit or harm
(AEs and harmful
outcomes) /
appreciable harm
(effectiveness and
beneficial outcomes)
Appreciable harm (AEs and
harmful outcomes) /
appreciable benefit
(effectiveness and beneficial
outcomes)
MID = minimal important difference determined for each outcome. The MIDs are the
threshold for appreciable benefits and harms. The confidence intervals of the top five points
of the diagram (within the green sector or within the purple sector) are considered precise
because the upper and lower limits of the point estimate (diamond shapes) do not cross the
pre-defined MID. Conversely, the bottom three points of the diagram are considered
imprecise because the upper and lower limits of the point estimates (diamonds) for each of
them cross the pre-defined MID and reduce the certainty of the result.
The following are the MID for the outcomes in this guideline (as agreed by the GDG).
Table 6: MIDs for the outcomes used in this guidance
Outcome
Relative risk reduction
Mortality from any cause
10%
Stroke (ischaemic or haemorrhagic)
10%
Myocardial infarction (MI) (including, where reported, silent MI)
10%
Heart failure
10%
20
Update 2011
0.75
Hypertension (partial update)
2011 Methods
Outcome
Relative risk reduction
New onset diabetes
10%
Vascular procedures (including both coronary and carotid artery procedures)
10%
Angina requiring hospitalisation
10%
Health-related quality of life (to use what is reported by trials)
As defined in literature for
each specific QoL measure
Major adverse cardiac and cerebrovascular events (MAACE): fatal and nonfatal MI, fatal and non-fatal stroke, hospitalised angina, hospitalised heart
failure, revascularisation (and different composites of this outcome)
15%
Study drug withdrawal rates (surrogate for adverse effects of drug treatment
and for adherence
10%
Angioedema in black people of African and Caribbean descent
10%
Blood pressure
5 mmHg (mean difference,
continuous outcome)
Prognostic studies
All prognostic study designs were included for the prognostic questions. The quality of the
prognostic studies was assessed using the quality checklist in the NICE Guidelines Manual
April 2009. The main criteria considered in assessing study quality were:
 The study sample represents the population of interest with regard to key characteristics,
sufficient to limit potential bias to the results
 Loss to follow-up is unrelated to key characteristics (that is, the study data adequately
represent the sample), sufficient to limit potential bias
 The prognostic factor of interest is adequately measured in study participants, sufficient to
limit potential bias
 The outcome of interest is adequately measured in study participants, sufficient to limit
bias
 Important potential confounders are appropriately accounted for, limiting potential bias
with respect to the prognostic factor of interest
 The statistical analysis is appropriate for the design of the study, limiting potential for the
presentation of invalid results
The methodological flaws of the prognostic studies included in the guideline update, have
been summarised in tables within appendix F, in order to give an overview of the quality of
each individual study, since GRADE is not currently designed for prognostic studies. Odds
ratios, relative risks or hazard ratios, with their 95% confidence intervals, from multivariate
analyses were extracted from the papers. Data for selected outcomes has been summarised in
tables within the relevant review chapter. Full data for all the outcomes has been reported in
the evidence tables (see appendix F) for each individual prognostic study. Taking into
consideration the advice on prognostic reviews in the NICE guidelines manual, meta-analysis
was not undertaken for prognostic studies.
3.3 Evidence of cost-effectiveness
3.3.1
Evidence on cost-effectiveness related to the key clinical issues being addressed in the
guideline was sought. The health economist undertook:
 A systematic review of the economic literature
 New cost-effectiveness analysis in priority areas
Literature review
The Health Economist:
21
Hypertension (partial update)
2011 Methods
 Identified potentially relevant studies for each review question from the economic search
results by reviewing titles and abstracts – full papers were then obtained.
 Reviewed full papers against pre-specified inclusion / exclusion criteria to identify relevant
studies (see below for details).
 Critically appraised relevant studies using the economic evaluations checklist as specified
in The Guidelines Manual.430
 Extracted key information about the study’s methods and results into evidence tables
(evidence tables are included in Appendix G: Evidence tables – health economic studies.
 Generated summaries of the evidence in NICE economic evidence profiles (included in the
relevant chapter write-ups) – see below for details.
Inclusion/exclusion
Full economic evaluations (studies comparing costs and health consequences of alternative
courses of action: cost–utility, cost-effectiveness, cost-benefit and cost-consequence analyses)
and comparative costing studies that addressed the review question in the relevant population
were considered potentially applicable as economic evidence.
Studies were excluded if they only reported cost per hospital (not per patient), or only
reported average cost effectiveness without disaggregated costs and effects. Abstracts,
posters, reviews, letters/editorials, foreign language publications and unpublished studies
were excluded. Studies judged to have an applicability rating of ‘not applicable’ were
excluded (this included studies that took the perspective of a non-OECD country).
Remaining studies were prioritised for inclusion based on their relative applicability to the
development of this guideline and the study limitations. For example, if a high quality,
directly applicable UK analysis was available other less relevant studies may have been
excluded and this is noted in the relevant section.
For more details about the assessment of applicability and methodological quality see the
economic evaluation checklist (The Guidelines Manual, Appendix H 430 and the health
economics research protocol in Appendix E: Review protocols.
When no relevant economic analyses were identified in the economic literature review,
relevant UK NHS unit costs were presented to the GDG to inform consideration of cost
effectiveness.
NICE economic evidence profiles
The NICE economic evidence profile has been used to summarise cost and cost-effectiveness
estimates. The economic evidence profile shows, for each economic study, an assessment of
applicability and methodological quality, with footnotes indicating the reasons for the
assessment. These assessments were made by the health economist using the economic
evaluation checklist from The Guidelines Manual, Appendix H.430 It also shows incremental
costs, incremental outcomes (for example, QALYs) and the incremental cost-effectiveness
ratio from the primary analysis, as well as information about the assessment of uncertainty in
the analysis. See Table 7 for more details.
If a non-UK study was included in the profile, the results were converted into pounds sterling
using the appropriate purchasing power parity.468
Table 7: Content of NICE economic profile
Item
Description
Study
First author name, reference, date of study publication and country perspective.
Limitations
An assessment of methodological quality of the study(a):
 Minor limitations – the study meets all quality criteria, or the study fails to meet
one or more quality criteria, but this is unlikely to change the conclusions about
Update 20
22
Hypertension (partial update)
2011 Methods
Item
Description
cost effectiveness.
 Potentially serious limitations – the study fails to meet one or more quality criteria,
and this could change the conclusion about cost effectiveness
 Very serious limitations – the study fails to meet one or more quality criteria and
this is very likely to change the conclusions about cost effectiveness. Studies with
very serious limitations would usually be excluded from the economic profile table.
Applicability
An assessment of applicability of the study to the clinical guideline, the current NHS
situation and NICE decision-making(a):
 Directly applicable – the applicability criteria are met, or one or more criteria are
not met but this is not likely to change the conclusions about cost effectiveness.
 Partially applicable – one or more of the applicability criteria are not met, and this
might possibly change the conclusions about cost effectiveness.
 Not applicable – one or more of the applicability criteria are not met, and this is
likely to change the conclusions about cost effectiveness.
Other comments
Particular issues that should be considered when interpreting the study.
Incremental cost
The mean cost associated with one strategy minus the mean cost of a comparator
strategy.
Incremental effects
The mean QALYs (or other selected measure of health outcome) associated with one
strategy minus the mean QALYs of a comparator strategy.
ICER
Incremental cost-effectiveness ratio: the incremental cost divided by the respective
QALYs gained.
Uncertainty
A summary of the extent of uncertainty about the ICER reflecting the results of
deterministic or probabilistic sensitivity analyses, or stochastic analyses of trial data,
as appropriate.
a) Limitations and applicability were assessed using the economic evaluation checklist from The Guidelines Manual,
Appendix H430
3.3.2
23
Update 2011
3.3.3
Undertaking new health economic analysis
As well as reviewing the published economic literature for each review question, as described
above, new cost-effectiveness analysis was undertaken by the Health Economist in priority
areas. Priority areas were agreed by the GDG after formation of the review questions and
consideration of the available health economic evidence.
Additional data for the analysis were identified as required through additional literature
searches undertaken by the Health Economist, and discussion with the GDG. Model structure,
inputs and assumptions were explained to and agreed by the GDG members during meetings,
and they commented on subsequent revisions. Results were presented in GDG meetings for
discussion and interpretation.
The priority area identified for new economic analysis was diagnosis of hypertension – see
‘Appendix J: Cost-effectiveness analysis – blood pressure monitoring for confirming a
diagnosis of hypertension (new 2011)’ for full methods. The 2006 cost-effectiveness analysis
of drug treatment was also updated – see ‘Appendix I: Cost-effectiveness analysis –
pharmacological treatment (updated 2011)’ for full methods.
Cost-effectiveness criteria
NICE’s report ‘Social value judgements: principles for the development of NICE guidance’
sets out the principles that GDGs should consider when judging whether an intervention
offers good value for money.429,430
In general, an intervention was considered to be cost effective if either of the following
criteria applied (given that the estimate was considered plausible):
a) The intervention dominated other relevant strategies (that is, it was both less costly in
terms of resource use and more clinically effective compared with all the other relevant
alternative strategies), or
Hypertension (partial update)
2011 Methods
b) The intervention cost less than £20,000 per quality-adjusted life-year (QALY) gained
compared with the next best strategy.
If the GDG recommended an intervention that was estimated to cost more than £20,000 per
QALY gained, or did not recommend one that was estimated to cost less than £20,000 per
QALY gained, the reasons for this decision are discussed explicitly in the ‘from evidence to
recommendations’ section of the relevant chapter with reference to issues regarding the
plausibility of the estimate or to the factors set out in the ‘Social value judgements: principles
for the development of NICE guidance’.429
3.4 Developing recommendations
3.4.1
3.4.3
3.4.4
24
Update 2011
3.4.2
Over the course of the guideline development process, the GDG was presented with:
 Evidence tables of the clinical and economic evidence reviewed from the literature. All
evidence tables are in Appendix E: Evidence Tables – Clinical studies and Appendix
G:Evidence tables – health economic studies.
 Summary of clinical and economic evidence and quality
 Forest plots and summary ROC curves
 A description of the methods and results of the cost-effectiveness analysis undertaken for
the guideline
The main considerations specific to each recommendation are outlined in the link from
evidence to recommendation section preceding the recommendation section.
Research recommendations
When areas were identified for which good evidence was lacking, the guideline development
group considered making recommendations for future research. Decisions about inclusion
were based on factors such as:
 the importance to patients or the population
 national priorities
 potential impact on the NHS and future NICE guidance
 ethical and technical feasibility
Validation process
The guidance is subject to a four week public consultation and feedback as part of the quality
assurance and peer review the document. All comments received from registered stakeholders
are responded to in turn and posted on the NICE website when the pre-publication check of
the full guideline occurs.
Updating the guideline
Following publication, and in accordance with the NICE guidelines manual, NICE will ask a
National Collaborating Centre or the National Clinical Guideline Centre to advise NICE’s
Guidance executive whether the evidence base has progressed significantly to alter the
guideline recommendations and warrant an update.
Disclaimer
Health care providers need to use clinical judgement, knowledge and expertise when deciding
whether it is appropriate to apply guidelines. The recommendations cited here are a guide and
may not be appropriate for use in all situations. The decision to adopt any of the
recommendations cited here must be made by the practitioners in light of individual patient
circumstances, the wishes of the patient, clinical expertise and resources.
The National Clinical Guideline Centre disclaims any responsibility for damages arising out
of the use or non-use of these guidelines and the literature used in support of these guidelines.
Hypertension (partial update)
2004 Methods
4
4.1.1
Funding
The National Clinical Guideline Centre was commissioned by the National Institute for
Health and Clinical Excellence to undertake the work on this guideline.
2004 Methods
Review methods
The aim of reviewing was to identify and synthesise relevant published and unpublished
evidence to allow recommendations to be evidence-based wherever possible.630 The search
was carried out using the electronic databases MEDLINE, EMBASE and CENTRAL,
attempting to locate systematic reviews and meta-analyses, and original randomised trials
using a combination of subject heading and free text searches. We made extensive use of high
quality recent review articles and bibliographies, as well as contact with subject area experts.
New searches were concentrated in areas of importance to the guideline development process,
for which existing systematic reviews were unable to provide valid or up to date answers. The
expert knowledge and experience of group members also backed up the search of the
literature.
Electronic searches used a sensitive search strategy based on a combination of text and index
terms to locate randomised controlled trials of treatments relevant to the guideline. If data
necessary for our analyses were not reported, we wrote to authors or sponsoring agencies. We
are grateful to investigators and sponsors who provided unpublished information to aid our
work.
We assessed the quality of relevant studies retrieved and their ability to provide valid answers
to the clinical questions addressed by the group. Assessment of study quality concentrated on
internal validity (the extent to which the study measured what it intended to measure),
external validity (the extent to which study findings could be generalised to other treatment
settings) and construct validity (the extent to which measurement corresponded to theoretical
understanding of a disease). 139
Table 8: Quality Criteria for Randomised Controlled Trials
Appropriateness of inclusion and exclusion criteria
Concealment of allocation
Blinding of patients
Blinding of health professionals
Blinding of data collectors/outcome assessors
Completeness and length of follow up
Appropriateness of outcome measures
Once data had been abstracted from individual papers and their quality assessed, the
information was synthesised. Individual trials often have an insufficient sample size to
identify significant outcomes with confidence81, so where appropriate, the results of
randomised studies were combined using meta-analytic techniques 175. Questions were
answered using the best evidence available. When considering the effect of an intervention, if
this could be addressed by the best study design then weaker designs were not reviewed.
Where studies were of poor quality, or contained patient groups considered likely to have
different responses, the effects of inclusion or exclusion were examined in sensitivity
analyses. No trials that met our inclusion criteria were excluded from the primary analyses.
However, where data on relevant outcomes were not available, these studies could not be
included, thus leading to the potential for publication bias.
25
Update 2011
3.4.5
Hypertension (partial update)
2004 Methods
Review criteria
Scoping work revealed a vast number of trials of pharmaceutical interventions. Recent work
suggests that study size is a useful proxy for study quality.189,224 Consequently to achieve the
task in the timescale provided we reviewed only those pharmaceutical studies which enrolled
200 or more patients. Since the prime motivation for treatment in hypertension, an
asymptomatic condition, is the prevention of mortality and morbidity, we reviewed those
studies with a planned follow-up of at least a year since such studies are likely to have been
designed to inform about these endpoints. Few non-pharmacological studies directly address
cardiovascular endpoints or feature substantial durations of follow-up. Consequently in these
areas we evaluated blood pressure reduction as a proxy endpoint and included trials with a
follow-up of 8 weeks follow-up or more, which compared a group receiving a lifestyle
intervention with a control group who received no treatment, usual treatment, sham therapy or
a placebo.
Statistical methods
Pharmacological interventions
The outcomes analyzed were: all cause mortality, fatal and non-fatal myocardial infarction,
fatal and non-fatal stroke. We did not consider the following endpoints: renal disease (rare in
non-diabetic patients); heart failure (inconsistently reported in trials); cardiovascular events (a
concatenation of myocardial infarction and stroke). For each trial, the risk ratios comparing
the risk of each outcome in the active treatment and control groups - or, for head-to-head
trials, in the different treatment groups - were calculated. Results of trials were combined in a
meta-analysis using the DerSimonian and Laird random effects model175, to estimate an
overall pooled risk ratio (RR) and its 95% confidence interval (95%CI). This model assumes
that there are different effects of treatment in different populations, which are clustered about
a mean effect; the pooled RR gives the best estimate of this mean effect. In the placebocontrolled trials reported in this guideline, a RR less than 1 favours treatment and a RR
greater than 1 favours control. If the 95%CI include 1, there is no statistically significant
difference between the treatments being compared.
Finally, we assessed the tolerability of the interventions by comparing the rate of overall
withdrawal (percentage of patients who withdrew each year) in each treatment arm of a trial
and calculating the difference in these rates (called the 'incident risk difference'). These
incident risk differences were combined in a meta-analysis using the DerSimonian and Laird
random effects model175, to estimate an overall pooled incident risk difference and its 95%
confidence interval.
We assessed heterogeneity between trials using a chi-squared statistic (Q). This assesses
whether the trials are sufficiently similar to be validly combined. Although the test for
heterogeneity is weak, it is usually assumed that if it gives p-values greater than 0.10, there is
no significant heterogeneity and it is valid to discuss the combined findings.
We also assessed whether the effect in individual trials was related to the size of the trial; any
such trend might indicate publication bias, e.g. where small trials were published only if they
showed a positive effect. Again, this test for systematic variation in the magnitude of the
estimated effect with the size of the trial is weak, but it is usually assumed that if it gives a pvalue greater than 0.10, there is unlikely to be any such bias.
Lifestyle interventions
None of the studies identified were designed to quantify significant changes in rates of death
or cardiovascular events, so we analysed the surrogate endpoint of reduced blood pressure.
For each trial, the difference in the final value mean blood pressure in the treatment and
control groups - or, for head-to-head trials, in the different treatment groups - was calculated.
26
Hypertension (partial update)
2004 Methods
4.1.2
4.1.3
Change scores from baseline were used where complete data for final values was unavailable.
These mean differences were weighted according to the precision of each trial (which depends
largely on its size, with larger trials getting more weight) and combined in a meta-analysis
using the DerSimonian and Laird random effects model175, to estimate an overall pooled
weighted mean difference and its 95% confidence interval. While most of the trials were of
parallel design (two or more groups received the various interventions at the same time),
some were of crossover design (all participants received both active treatment and control
interventions, but in a random order). Crossover trials have about four times greater precision
than parallel trials of the same size, so we used methods have been developed recently to
combine the parallel and crossover trials in the same meta-analysis.147,193 Heterogeneity and
the potential for publication bias were assessed in the same way as for pharmaceutical trials.
The mean percentage achieving a reduction of 10mmHg or more in systolic blood pressure
was then estimated from the cumulative normal distribution637 and confidence intervals were
estimated using the delta method.51
Finally, we assessed the tolerability of the interventions by comparing the proportion of
withdrawals (% of patients who withdrew) in each treatment arm of a trial and calculating the
difference in these proportion (called the 'risk difference'). These risk differences were
combined in a meta-analysis using the DerSimonian and Laird random effects model,175 to
estimate an overall pooled risk difference and its 95% confidence interval.
Group process
The guideline development group was run using the principles of small group work and was
led by a trained facilitator. The group underwent initial exercises to set its own rules to
determine how it wanted to function and received brief training on reviewing methods,
economic analysis and grading methodology. Additional training was provided in the group
as the need arose in subsequent meetings. Findings, expressed as narratives, statements of
evidence and recommendations, were reached by informal consensus. There was no
obligation to force an agreement where none existed after discussion: dissensions were
recorded in the guideline narrative.471
Evidence statements and recommendations
The guideline development group process produces summary statements of the evidence
concerning available treatments and healthcare and from these makes its recommendations.
Evidence statements and recommendations are commonly graded in guidelines reflecting the
quality of the study designs on which they are based. An established scheme adapted from the
Agency for Health Care Policy and Research (AHCPR) Classification is shown in Table 9 and
Table 10.14
Table 9: AHCPR derived categories of evidence
Level of evidence
Ia:
evidence from meta-analysis of randomised controlled trials
Ib:
evidence from at least one randomised controlled trial
IIa:
evidence from at least one controlled study without randomisation
IIb
:
evidence from at least one other type of quasi-experimental study
III:
evidence from non-experimental descriptive studies, such as comparative studies, correlation studies
and case-control studies
IV:
evidence from expert committee reports or opinions and/or clinical experience of respected authorities
Table 10: AHCPR derived strengths of recommendations
Strength of evidence
A
directly based on category I evidence
27
Hypertension (partial update)
2004 Methods
Strength of evidence
B
directly based on category II evidence or extrapolated recommendation from category I evidence
C
directly based on category III evidence or extrapolated recommendation from category I or II evidence
D
directly based on category IV evidence or extrapolated recommendation from category I, II or III
evidence
Two grading schemes were used when developing this guideline, the one above and a new
scheme called GREG (Guideline Recommendation and Evidence Grading).392 The new
scheme seeks to address a number of problems, by extending grading from treatment to
include diagnosis, prognosis and cost, and to handle the subtleties of clinical evidence more
sensitively (Table 11).
Table 11: GREG scheme for assessing evidence and writing recommendations
EVIDENCE
Evidence statements provide information about disease, diagnosis and treatment, and are used to
support recommendations. Each evidence statement is graded by scoring the study design and applying
quality corrections.
Design
Notes
Design scores
Treatment
Randomised controlled trial
Non-randomised controlled study
Uncontrolled study
1
2
3
Diagnosis
Blinded cohort study
Unblinded cohort study
Other design
1
2
3
Prognosis
Incidence cohort study
Other cohort study
Descriptive data
Population data
Representative sample
Convenience sample
1
2
3
1
2
3
Notes
i. Blinding refers to independent interpretation of a
test and reference standard.
ii. An incident cohort is identified and followed in
time from a defined point in the progress of
disease or care.
iii. Important flaws may be judged to occur when
adequate standards of research are not followed
or are unreported in published findings. Potential
examples include failure to analyse by intentionto-treat, over-interpretation of secondary
analyses, failure to adjust for potential
confounding in non- randomised designs. For
diagnostic studies this includes the need for an
adequate reference standard and to apply
different tests in an adequately short timescale.
iv. Sparse data (too few events or patients) are the
most common reason for imprecision. A
confidence interval including both no effect and a
clinically important effect is an example of an
imprecise finding.
v. Consistency in [1] design: involves methods,
patients, outcome measures; and [2] findings:
involves homogeneity of summary estimates.
Independence refers to the availability of research
from at least two independent sources. Evidence
of publication bias also denotes lack of
consistency.
vi. Adequate relevance requires [1] use in studies of
a relevant patient-oriented health outcome or a
strongly linked surrogate endpoint; and [2] a
sufficiently representative and relevant patient
group or mix.
vii. In comparative designs a very strong association
can raise the quality score.
Quality corrections
Flawed design, conduct or analysis
Imprecise findings
Lack of consistency or independence
Inadequate relevance
Very strong association
Evidence Grade
I: High
II: Intermediate 2
III: Low
+1
+1
+1
+1
+1
-1
≤1
2
≥3
Recommendations
Recommendations provide guidance about appropriate care. Ideally, these should be based on clear evidence:
a robust understanding of the benefits, tolerability, harms and costs of alternative patterns of care. They also
28
Hypertension (partial update)
2004 Methods
EVIDENCE
Evidence statements provide information about disease, diagnosis and treatment, and are used to
support recommendations. Each evidence statement is graded by scoring the study design and applying
quality corrections.
need to be feasible in the healthcare setting addressed. There are three unique categories, and each
recommendation may be positive or negative, conditional or unconditional reflecting current evidence and the
understanding of the guideline group.
A. Recommendation – There is robust evidence to recommend a pattern of care.
B. Provisional recommendation – On balance of evidence, a pattern of care is recommended with
caution.
C. Consensus Opinion – Evidence being inadequate, a pattern of care is recommended by consensus.
4.1.4
Use of the two schemes was evaluated in this and another guideline being developed
contemporaneously. Both groups consistently favoured the new scheme and so the guideline
is presented using the new grading scheme. The evaluation of the two schemes will be
reported separately.
The key point of note is that any assessment of evidence quality is ultimately a subjective
process. How bad does a trial have to be before it is flawed or how sparse do the findings
have to be before we lose confidence in the findings? The purpose of an evidence grading
scheme is to characterise the robustness of outcomes from studies, and the random and
systematic biases that pertain to them.
Similarly recommendation grading must credibly assimilate evidence and health service
context to credibly advise lines of care for average patients. Clinicians must use their
judgement and awareness of patients' circumstances and values when considering
recommendations from guidelines.
Costs and consequences
Approaches to cost-effectiveness have assisted in reaching recommendations in a series of
primary care evidence-based guidelines.188,393 This guideline involves a systematic appraisal
of effectiveness, compliance, quality-of-life, safety and health service resource use and costs
of a medical intervention provided in the British health care setting. Using the most current,
pertinent and complete data available, the economic analysis attempts a robust presentation
showing the possible bounds of cost-effectiveness that may result.
The guiding principle behind economic analysis is that it is desirable to use limited healthcare
resources to maximise health improvements in the population. Well defined but narrow
notions of health improvement may not reflect all aspects of value to patients, carers,
clinicians or society. For example, evidence may lead the guideline group to recommend
targeting additional resources to certain patient groups when unequal access to care is
apparent. The group process allows discussion of what should be included in the definition of
'improved health' and more broadly of other concepts of value to society such as fairness,
justice, dignity or minimum standards of care.
 The range of values used to generate cost-effectiveness estimates reflects the available
evidence and the concerns of the guideline development group. Recommendations are
graded reflecting the certainty with which the costs and consequences of a medical
intervention can be assessed. This practice reflects the desire of group members to have
simple, understandable and robust information based on good data.
 It is not generally helpful to present an additional systematic review of previous economic
analyses that have adopted a variety of differing perspectives, analytic techniques and
baseline data. However, the economic literature is reviewed to compare guideline findings
with representative published economic analyses and to interpret any differences in
findings when these occurred. A commentary is included when the group feel this aids
understanding.
29
Hypertension (partial update)
2004 Methods
2006 methods
Clinical evidence
Methodological introduction
Study inclusion and reporting criteria
A systematic search of the literature was performed on EMBASE and MEDLINE for
randomised controlled trials comparing any combination of antihypertensive drugs from
among the following five classes of drugs:
 ACE inhibitors (ACEi)
 angiotensin-II receptor antagonists (ARB)
 beta-receptor blockers (BB)
 calcium-channel blockers (CCB)
 thiazide-type diuretics (TD).
Placebo-controlled studies were not included because the main aim of this rapid partial update
was to make recommendations regarding the optimal sequencing of drug treatment for
hypertension, for which head-to-head studies are required, and because sufficient placebocontrolled studies of the main drug classes had been considered in the original NICE
guideline. However, placebo-controlled studies were sought for isolated systolic hypertension
because of a lack of comparator studies.
The cut-off date for evidence to be considered in the previous guideline was July 2004, so this
update only searched for English-language titles published after that date. Papers published up
to and including 19 December 2005 were considered – this constitutes the cut-off for evidence
for this rapid update.
Studies were excluded due to:
 inadequate or no randomisation
 inadequate study power, defined as a sample size of less than 200 patients, or having a
follow-up period of less than 12 months
 having an exclusive diabetic or paediatric patient population, unrepresentative of the
general UK hypertensive population
 stroke, myocardial infarction, and mortality outcomes not being reported.
The following outcomes were recorded for each study, where available:
 mortality from any cause
 stroke (ischaemic or haemorrhagic)
 myocardial infarction (including, where reported, silent MI)
 heart failure
 new-onset diabetes mellitus
 vascular procedures (including both coronary and carotid artery procedures)
 incidence of unstable angina (or angina episodes requiring hospitalisation)
 study drug withdrawal.
Interpretation and analysis of results
All outcomes, with the exception of study drug withdrawal, vascular procedures and unstable
angina, were entered into a meta-analysis for each drug combination using RevMan 4.2
software (©The Nordic Cochrane Centre). The overall effect size was reported as the relative
risk (RR) with 95% confidence intervals in each case.
30
Hypertension (partial update)
2004 Methods
Update
2011
A p-value less than 0.05 was considered statistically significant for overall effect. Forest plots
for each comparison are included in Appendix A.
In recording the outcomes, stroke was considered to be synonymous with 'cerebrovascular
event'. Reports of 'cardiovascular events' or other composite outcomes other than those listed
above were not considered.
Sensitivity analyses were performed based on the inclusion and exclusion of silent myocardial
infarction and the inclusion and exclusion of secondary prevention studies. Additional
subgroup analyses were performed to identify the source of any significant heterogeneity in
study results (defined as an I2 statistic greater than 50%).
Where the heterogeneity has I2 greater than 50%, the trials are reported individually in the
evidence statements.
The following outcomes were not subject to meta-analysis due to potential variability or
subjectivity in diagnosis or treatment protocols, and were reported as a narrative only:
 unstable angina
 revascularisation procedures
 study drug withdrawal.
Following consultation on the draft guideline, heart failure as an outcome was included in the
meta-analysis. Because of inconsistency in definition of heart failure in the trials, this was
analysed using a random effects model.
Secondary analyses
In addition to results in general hypertensive populations, the following subgroups were also
considered separately:
 those patients with isolated systolic hypertension (ISH)
 black people of African and Caribbean descent younger patients (defined as under 55
years).
For ISH, due to the lack of evidence comparing different antihypertensive drugs, the results
from placebo-controlled trials were also considered. These results included pre-defined
subgroup analyses from trials in general hypertensive populations as well as one trial
comprising only ISH patients. The results were entered into a meta-analysis according to the
same procedure specified above. The definition of ISH varied slightly between studies:
permitting a diastolic blood pressure up to 95 mmHg in one study (SYST-EUR43,124,555) and
90 mmHg in the others (SHEP483,536,537,606, SHEP-P281,484,485).
No trials comprising only non-white patients were found, although two pre-defined subgroup
analyses from trials in general hypertensive populations were found (ALLHAT589-591,
LIFE154,176,222,369,370,507,618,619). Results involving placebo comparisons in non-white
populations were not considered.
Evidence on younger patients was extremely sparse, and evidence consideration was therefore
extended to include papers pre-dating July 2004 and in which blood pressure lowering effect
was the main outcome measure.
Cost-effectiveness evidence
The GDG drafted recommendations on the basis of the clinical evidence. A health economic
analysis was then conducted to balance the clinical outcomes and to test the cost effectiveness
of different initial antihypertensive medications.
See ‘Appendix I: Cost-effectiveness analysis – pharmacological treatment (updated 2011)’ for
full methods – note that analysis was updated as part of the 2011 update.
31
Hypertension (partial update)
Guideline summary
5
Guideline summary
5.1 Algorithms
Figure 2: Diagnosis of Hypertension
Update 2011
32
Hypertension (partial update)
Guideline summary
Figure 3: Treatment of Hypertension
Update 2011
33
Hypertension (partial update)
Guideline summary
5.2 Key priorities for implementation
34
Updte 2011
From the full set of recommendations, the GDG selected ten key priorities for
implementation. The criteria used for selecting these recommendations are listed in detail in
The Guidelines Manual.430 The reasons that each of these recommendations was chosen are
shown in the table linking the evidence to the recommendation in the relevant chapter.
Diagnosing hypertension
 If the clinic blood pressure is 140/90 mmHg or higher, offer ambulatory blood pressure
monitoring (ABPM) to confirm the diagnosis of hypertension. [new 2011]
 When using ABPM to confirm a diagnosis of hypertension, ensure that at least two
measurements per hour are taken during the person’s usual waking hours (for example,
between 08:00 and 22:00).
Use the average value of at least 14 measurements taken during the person’s usual waking
hours to confirm a diagnosis of hypertension. [new 2011]
 When using home blood pressure monitoring (HBPM) to confirm a diagnosis of
hypertension, ensure that:
o for each blood pressure recording, two consecutive measurements are taken, at least 1
minute apart and with the person seated and
o blood pressure is recorded twice daily, ideally in the morning and evening and
o blood pressure recording continues for at least 4 days, ideally for 7 days.
Discard the measurements taken on the first day and use the average value of all the
remaining measurements to confirm a diagnosis of hypertension. [new 2011]
Initiating treatment
 Offer antihypertensive drug treatment to people aged under 80 years with stage 1
hypertension who have one or more of the following:
o target organ damage
o established cardiovascular disease
o renal disease
o diabetes
o a 10-year cardiovascular risk equivalent to 20% or greater. [new 2011]
 Offer antihypertensive drug treatment to people of any age with stage 2 hypertension. [new
2011]
 For people aged under 40 years with stage 1 hypertension and no evidence of target organ
damage, cardiovascular disease, renal disease or diabetes, consider seeking specialist
evaluation of secondary causes of hypertension and a more detailed assessment of potential
target organ damage. This is because 10-year cardiovascular risk assessments can
underestimate the lifetime risk of cardiovascular events in these people. [new 2011]
Monitoring treatment and blood pressure targets
 For people identified as having a ‘white-coat effect’, consider ABPM or HBPM as an
adjunct to clinic blood pressure measurements to monitor the response to antihypertensive
treatment with lifestyle modification or drugs. [new 2011]
Choosing antihypertensive drug treatment
 Offer people aged 80 years and over the same antihypertensive drug treatment as people
aged 55–80 years, taking into account any comorbidities. [new 2011]
Step 1 treatment
Hypertension (partial update)
Guideline summary
 Offer step 1 antihypertensive treatment with a calcium-channel blocker (CCB) to people
aged over 55 years and to black people of African or Caribbean family origin of any age. If
a CCB is not suitable, for example because of oedema or intolerance, or if there is evidence
of heart failure or a high risk of heart failure, offer a thiazide-like diuretic. [new 2011]
 If diuretic treatment is to be initiated or changed, offer a thiazide-like diuretic, such as
chlortalidone (12.5–25.0 mg once daily) or indapamide (1.5 mg modified-release or 2.5 mg
once daily) in preference to a conventional thiazide diuretic such as bendroflumethiazide or
hydrochlorothiazide. [new 2011]
 For people who are already having treatment with bendroflumethiazide or
hydrochlorothiazide and whose blood pressure is stable and well controlled, continue
treatment with the bendroflumethiazide or hydrochlorothiazide. [new 2011]
Step 4 treatment
 For treatment of resistant hypertension at step 4:
o Consider further diuretic therapy with low-dose spironolactone (25 mg once daily) if the
blood potassium level is 4.5 mmol/l or lower. Use particular caution in people with a
reduced estimated glomerular filtration rate because they have an increased risk of
hyperkalaemia.
o Consider higher-dose thiazide-like diuretic treatment if the blood potassium level is
higher than 4.5 mmol/l. [new 2011]
Full list of recommendations
1. Healthcare professionals taking blood pressure measurements need adequate initial training
and periodic review of their performance. [2004]
2. Because automated devices may not measure blood pressure accurately if there is pulse
irregularity (for example, due to atrial fibrillation), palpate the radial or brachial pulse
before measuring blood pressure. If pulse irregularity is present, measure blood pressure
manually using direct auscultation over the brachial artery. [new 2011]
3. Healthcare providers must ensure that devices for measuring blood pressure are properly
validated, maintained and regularly recalibrated according to manufacturers’ instructions.
[2004]
4. When measuring blood pressure in the clinic or in the home, standardise the environment
and provide a relaxed, temperate setting, with the person quiet and seated, and their arm
outstretched and supported. [new 2011]
5. If using an automated blood pressure monitoring device, ensure that the device is validated
and an appropriate cuff size for the person’s arm is used. [new 2011]
6. In people with symptoms of postural hypotension (falls or postural dizziness):
 measure blood pressure with the person either supine or seated.
 measure blood pressure again with the person standing for at least a minute prior to
measurement. [2004, amended 2011]
7. If the systolic blood pressure falls by 20 mmHg or more when the person is standing:
 review medication
 measure subsequent blood pressures with the person standing
 consider referral to specialist care if symptoms of postural hypotension persist. [2004,
amended 2011]
8. When considering a diagnosis of hypertension, measure blood pressure in both arms:
 If the difference in readings between arms is more than 20 mmHg, repeat the
measurements.
35
Hypertension (partial update)
Guideline summary
36
Update 2011
 If the difference in readings between arms remains more than 20 mmHg on the second
measurement, measure subsequent blood pressure in the arm with the higher reading.
[new 2011]
9. If blood pressure measured in the clinic is 140/90mmHg or higher:
 Take a second measurement during consultation.
 If the second measurement is substantially different from the first, take a third
measurement.
Record the lower of the last two measurements as the clinic blood pressure. [new 2011]
10.
If the clinic blood pressure is 140/90 mmHg or higher, offer ambulatory blood
pressure monitoring (ABPM) to confirm the diagnosis of hypertension. [new 2011]
11.
If a person is unable to tolerate ABPM, home blood pressure monitoring (HBPM) is a
suitable alternative to confirm the diagnosis of hypertension. [new 2011]
12.
If the person has severe hypertension, consider starting antihypertensive drug
treatment immediately, without waiting for the results of ABPM or HBPM. [new 2011]
13.
While waiting for confirmation of a diagnosis of hypertension, carry out investigations
for target organ damage (such as left ventricular hypertrophy, chronic kidney disease and
hypertensive retinopathy) (see 21) and a formal assessment of cardiovascular risk using a
cardiovascular risk assessment tool (see 20). [new 2011]
14.
If hypertension is not diagnosed but there is evidence of target organ damage such as
left ventricular hypertrophy, albuminuria or proteinuria, consider carrying out
investigations for alternative causes of the target organ damage. [new 2011]
15.
If hypertension is not diagnosed, measure the person’s clinic blood pressure at least
every 5 years subsequently, and consider measuring it more frequently if the person’s
clinic blood pressure is close to 140/90 mmHg. [new 2011]
16.
When using ABPM to confirm a diagnosis of hypertension, ensure that at least two
measurements per hour are taken during the person’s usual waking hours (for example,
between 08:00 and 22:00). Use the average value of at least 14 measurements taken during
the person’s usual waking hours to confirm a diagnosis of hypertension. [new 2011]
17.
When using HBPM to confirm a diagnosis of hypertension, ensure that:
 for each blood pressure recording, two consecutive measurements are taken, at least 1
minute apart and with the person seated and
 blood pressure is recorded twice daily, ideally in the morning and evening and
 blood pressure recording continues for at least 4 days, ideally for 7 days.
Discard the measurements taken on the first day and use the average value of all the
remaining measurements to confirm a diagnosis of hypertension. [new 2011]
18.
Refer the person to specialist care the same day if they have:
 accelerated hypertension, that is, blood pressure usually higher than 180/110 mmHg
with signs of papilloedema and/or retinal haemorrhage or
 suspected phaeochromocytoma (labile or postural hypotension, headache, palpitations,
pallor and diaphoresis). [2004, amended 2011]
19.
Consider the need for specialist investigations in people with signs and symptoms
suggesting a secondary cause of hypertension. [2004, amended 2011]
For NICE guidance on the early identification and management of chronic kidney disease see
‘Chronic kidney disease’ (NICE clinical guideline 73, 2008).
20.
Use a formal estimation of cardiovascular risk to discuss prognosis and healthcare
options with people with hypertension, both for raised blood pressure and other modifiable
risk factors. [2004]
21.
Estimate cardiovascular risk in line with the recommendations on Identification and
assessment of CVD risk in ‘Lipid modification’ (NICE clinical guideline 67). [2008]
Hypertension (partial update)
Guideline summary
37
Update 2011
22.
For all people with hypertension offer to:
 test for the presence of protein in the urine by sending a urine sample for estimation of
the albumin:creatinine ratio and test for haematuria using a reagent strip
 take a blood sample to measure plasma glucose, electrolytes, creatinine, estimated
glomerular filtration rate, serum total cholesterol and HDL cholesterol
 examine the fundi for the presence of hypertensive retinopathy
 arrange for a 12-lead electrocardiograph to be performed. [2004, amended 2011]
23.
Offer antihypertensive drug treatment to people aged under 80 years with stage 1
hypertension who have one or more of the following:
 target organ damage
 established cardiovascular disease
 renal disease
 diabetes
 a 10-year cardiovascular risk equivalent to 20% or greater. [new 2011]
24.
Offer antihypertensive drug treatment to people of any age with stage 2 hypertension.
[new 2011]
25.
For people aged under 40 years with stage 1 hypertension and no evidence of target
organ damage, cardiovascular disease, renal disease or diabetes, consider seeking specialist
evaluation of secondary causes of hypertension and a more detailed assessment of potential
target organ damage. This is because 10-year cardiovascular risk assessments can
underestimate the lifetime risk of cardiovascular events in these people. [new 2011]
26.
Use clinic blood pressure measurements to monitor the response to antihypertensive
treatment with lifestyle modification or drugs. [new 2011]
27.
For people identified as having a ‘white-coat effect’, consider ABPM or HBPM as an
adjunct to clinic blood pressure measurements to monitor the response to antihypertensive
treatment with lifestyle modification or drugs. [new 2011]
28.
Aim for a target clinic blood pressure below 140/90 mmHg in people aged under 80
years with treated hypertension. [new 2011]
29.
Aim for a target clinic blood pressure below 150/90 mmHg in people aged 80 years
and over with treated hypertension. [new 2011]
30.
When using ABPM or HBPM to monitor the response to treatment (for example, in
people identified as having a ‘white-coat effect’ and people who choose to monitor their
blood pressure at home) aim for a target average blood pressure during the person’s usual
waking hours of:
 below 135/85 for people aged under 80 years
 below 145/85 in people aged over 80 years and over. [new 2011]
For NICE guidance on the prevention of obesity and cardiovascular disease see ‘Obesity’
(NICE clinical guideline 43, 2006) and ‘Prevention of cardiovascular disease at population
level’ (NICE public health guidance 25, 2010).
31.
Lifestyle advice should be offered initially and then periodically to people undergoing
assessment or treatment for hypertension. [2004]
32.
Ascertain people’s diet and exercise patterns because a healthy diet and regular
exercise can reduce blood pressure. Offer appropriate guidance and written or audiovisual
materials to promote lifestyle changes. [2004]
33.
Relaxation therapies can reduce blood pressure and people may wish to pursue these
as part of their treatment. However, routine provision by primary care teams is not
currently recommended. [2004]
Hypertension (partial update)
Guideline summary
38
Update 2011
34.
Ascertain people’s alcohol consumption and encourage a reduced intake if they drink
excessively, because this can reduce blood pressure and has broader health benefits. [2004]
35.
Discourage excessive consumption of coffee and other caffeine-rich products.
36.
Encourage people to keep their dietary sodium intake low, either by reducing or
substituting sodium salt, as this can reduce blood pressure.[2004]
37.
Do not offer calcium, magnesium or potassium supplements as a method for reducing
blood pressure. [2004]
38.
The best current evidence does not show that combinations of potassium, magnesium
and calcium supplements reduce blood pressure. [2004]
39.
Offer advice and help to smokers to stop smoking. [2004]
40.
A common aspect of studies for motivating lifestyle change is the use of group
working. Inform people about local initiatives by, for example, healthcare teams or patient
organisations that provide support and promote healthy lifestyle change. [2004]
41.
Where possible, recommend treatment with drugs taken only once a day. [2004]
42.
Prescribe non-proprietary drugs where these are appropriate and minimise cost. [2004]
43.
Offer people with isolated systolic hypertension (systolic BP 160 mmHg or more) the
same treatment as people with both raised systolic and diastolic blood pressure. [2004]
44.
Offer people aged 80 years and over the same antihypertensive drug treatment as
people aged 55–80 years, taking into account any comorbidities. [new 2011]
45.
Offer antihypertensive drug treatment to women of child-bearing potential in line with
the recommendations on Management of pregnancy with chronic hypertension and
Breastfeeding in ‘Hypertension in pregnancy’ (NICE clinical guideline 107). [2010]
46.
Offer people aged under 55 years step 1 antihypertensive treatment with an
angiotensin-converting enzyme (ACE) inhibitor or a low-cost angiotensin-II receptor
blocker (ARB). If an ACE inhibitor is prescribed and is not tolerated (for example, because
of cough), offer a low-cost ARB. [new 2011]
47.
Do not combine an ACE inhibitor with an ARB to treat hypertension. [new 2011]
48.
Offer step 1 antihypertensive treatment with a calcium-channel blocker (CCB) to
people aged over 55 years and to black people of African or Caribbean family origin of any
age. If a CCB is not suitable, for example because of oedema or intolerance, or if there is
evidence of heart failure or a high risk of heart failure, offer a thiazide-like diuretic. [new
2011]
49.
If a diuretic is to be initiated or changed, offer a thiazide-like diuretic, such as
chlortalidone (12.5 mg–25.0 mg once daily) or indapamide (1.5mg slow release or 2.5 mg
once daily) in preference to a conventional thiazide diuretic such as bendroflumethiazide or
hydrochlorothiazide. [new 2011]
50.
For people who are already having treatment with bendroflumethiazide or
hydrochlorothiazide and whose blood pressure is stable and well controlled, continue
treatment with the bendroflumethiazide or hydrochlorothiazide. [new 2011]
51.
Beta-blockers are not a preferred initial therapy for hypertension. However, betablockers may be considered in younger people, particularly:
 those with an intolerance or contraindication to ACE inhibitors and angiotensin-II
receptor antagonists or
 women of child-bearing potential or
 people with evidence of increased sympathetic drive. [2006]
52.
If therapy is initiated with a beta-blocker and a second drug is required, add a calciumchannel blocker rather than a thiazide-like diuretic to reduce the person’s risk of
developing diabetes. [2006]
53.
If blood pressure is not controlled by Step 1 treatment, offer step 2 treatment with a
CCB in combination with either an ACE inhibitor or an ARB. [new 2011]
Hypertension (partial update)
Guideline summary
5.3 Key research recommendations
1. Which automated blood pressure monitors are suitable for people with hypertension and
atrial fibrillation?
39
Update 2011
54.
If a CCB is not suitable for step 2 treatment, for example because of oedema or
intolerance, or if there is evidence of heart failure or a high risk of heart failure, offer a
thiazide-like diuretic. [new 2011]
55.
For black people of African or Caribbean family origin, consider an ARB in
preference to an ACE inhibitor, in combination with a CCB. [new 2011]
56.
Before considering step 3 treatment, review medication to ensure step 2 treatment is at
optimal or best tolerated doses. [new 2011]
57.
If treatment with three drugs is required, the combination of ACE inhibitor or
angiotensin-II receptor blocker, calcium-channel blocker and thiazide-like diuretic should
be used. [2006]
58.
Regard clinic blood pressure that remains higher than 140/90 mmHg after treatment
with the optimal or best tolerated doses of an ACE inhibitor or an ARB plus a CCB plus a
diuretic as resistant hypertension, and consider adding a fourth antihypertensive drug
and/or seeking expert advice. [new 2011]
59.
For treatment of resistant hypertension at step 4:
 Consider further diuretic therapy with low-dose spironolactone (25 mg once daily). If
the blood potassium level is 4.5 mmol/l or lower. Use particular caution in people with a
reduced estimated glomerular filtration rate because they have an increased risk of
hyperkaelemia.
 Consider higher-dose thiazide-like diuretic treatment if the blood potassium level is
higher than 4.5 mmol/l. [new 2011]
60.
When using further diuretic therapy for resistant hypertension at step 4, monitor blood
sodium and potassium and renal function within 1 month and repeat as required thereafter.
[new 2011]
61.
If further diuretic therapy for resistant hypertension at step 4 is not tolerated, or is
contraindicated or ineffective, consider an alpha- or beta-blocker. [new 2011]
62.
If blood pressure remains uncontrolled with the optimal or maximum tolerated doses
of four drugs, seek expert advice if it has not yet been obtained. [new 2011]
63.
Provide appropriate guidance and materials about the benefits of drugs and the
unwanted side effects sometimes experienced in order to help people make informed
choices. [2004]
64.
People vary in their attitudes to their hypertension and their experience of treatment. It
may be helpful to provide details of patient organisations that provide useful forums to
share views and information. [2004]
65.
Provide an annual review of care to monitor blood pressure, provide people with
support and discuss their lifestyle, symptoms and medication. [2004]
66.
Because evidence supporting interventions to increase adherence is inconclusive, only
use interventions to overcome practical problems associated with non-adherence if a
specific need is identified. Target the intervention to the need. Interventions might include:
 suggesting that patients record their medicine-taking
 encouraging patients to monitor their condition
 simplifying the dosing regimen
 using alternative packaging for the medicine using a multi-compartment medicines
system. (This recommendation is taken from ‘Medicines adherence’, NICE 408 clinical
guideline 76). [new 2011]
Hypertension (partial update)
Measuring blood pressure
Measuring blood pressure
For many years blood pressure has been measured using a brachial pressure cuff and
auscultation of the brachial artery to identify the appearance and disappearance of Korotkoff
sounds. Increasingly, automated devices for measuring blood pressure are now used in the
clinic, hospitals and by people in their homes. In addition, ambulatory blood pressure
measurement devices are available that are programmed to allow blood pressure to be
measured repeatedly during the day and night. Blood pressure (BP) can be highly variable and
this variability is due to the inherent variability in BP itself and the influence of factors such
as posture, room temperature and pain/discomfort or stress. In addition there are factors
related to the process of BP measurement itself that can contribute to BP variability such as
the appropriateness of the cuff size, the rate of inflation and deflation of the cuff and the
accuracy of the process of measurement or the automated BP monitor being used.
5.4 Techniques for measuring blood pressure
5.4.1
Manual blood pressure measurement
The cuff is inflated to block the brachial pulse. The first sound occurring with the return of the
brachial pulse is the systolic pressure (the point at which the heart pumping at its hardest
overcomes the pressure exerted by the cuff to push blood past the obstruction). Intermediate
sounds follow as the cuff pressure drops, with muffling and then the disappearance of sounds
indicating the diastolic pressure (the point at which the heart is not pumping outward and the
residual arterial pressure is sufficient to overcome the pressure exerted by the cuff). The
interpretation of the sounds was later developed by Ettinger.579
Three types of error have been identified for the RRK technique. Failure to accurately identify
the Korotkoff sounds can lead to over or under estimation. Digit preference refers to the
tendency of clinicians to round readings up or down, often to the nearest zero. Observer
prejudice occurs when clinicians alter readings toward their prior expectation, a particular
concern when close to a threshold which changes management.64,482 Supervised training and
reassessment may help minimise errors.
Systolic pressure is estimated by first palpating the brachial pulse with slow deflation of the
cuff. The cuff is reinflated before listening for Korotkoff sounds. The first pass is important
since sometimes the first sounds disappear as pressure is reduced (the auscultatory gap)
leading to an underestimation of systolic pressure by auscultation alone. In a case series, 21%
of 168 untreated hypertensive patients demonstrated an auscultatory gap.121 A number of
summaries are available highlighting good technique: an adaptation of these is shown in
Table 12.
Table 12: Estimating blood pressure by manual auscultation
Manual auscultation
Standardise the environment as much as possible:
40
Update 2011
2. In people aged under 40 with hypertension, what is the most accurate method of assessing
the lifetime risk of cardiovascular events and the impact of therapeutic intervention on this
risk?
3. In people aged under 40 with hypertension, what are the appropriate thresholds for
intervention?
4. In adults with primary hypertension, does the use of out-of-office monitoring (HBPM or
ABPM) improve response to treatment?
5. In people with treated hypertension, what is the optimal systolic blood pressure?
6. In adults with hypertension, which drug treatment (diuretic therapy versus other step 4
treatments) is the most clinically and cost effective for step 4 treatment?
Hypertension (partial update)
Measuring blood pressure
Manual auscultation
 Relaxed, temperate setting, with the patient seated and rested
 Arm out-stretched, in line with mid-sternum and supported
 Correctly wrap a cuff containing an appropriately sized bladder around the upper arm and connect to a
manometer. Cuffs should be marked to indicate the range of permissible arm circumferences; these marks
should be easily seen when the cuff is being applied to an arm.
 Palpate the brachial pulse in the antecubital fossa of that arm.
 Rapidly inflate the cuff to 20 mmHg above the point where the brachial pulse disappears.
 Deflate the cuff and note the pressure at which the pulse reappears: the approximate systolic pressure.
 Re-inflate the cuff to 20 mmHg above the point at which the brachial pulse disappears.
 Using one hand, place the stethoscope over the brachial artery ensuring complete skin contact with no
clothing in between.
 Slowly deflate the cuff at 2–3 mmHg per second listening for the Korotkoff sounds.
Phase I: The first appearance of faint repetitive clear tapping sounds gradually increasing in intensity and
lasting for at least two consecutive beats: note the systolic pressure.
Phase II: A brief period may follow when the sounds soften and or 'swish'.
Auscultatory Gap: In some patients the sounds may disappear altogether.
Phase III: The return of sharper sounds becoming crisper for a short time.
Phase IV: The distinct, abrupt muffling of sounds, becoming soft and blowing in quality.
Phase V: The point at which all sounds disappear completely: note the diastolic pressure.
 When the sounds have disappeared, quickly deflate the cuff completely if repeating the measurement.
 When possible, take readings at the beginning and end of consultations.
There has been some controversy as to whether phase IV or phase V sounds should be used to
record diastolic blood pressure. Commonly, the difference in pressure between phase IV and
V is less than 5 mmHg but occasionally can be substantial. Phase V can be absent with sounds
audible to zero cuff pressure notably in some children, during pregnancy, with anaemia, aortic
insufficiency and with elderly people. Phase V correlates better with direct measurement, is
commonly used in clinical trials of antihypertensive therapies, and is more reproducible when
assessed by different observers. There is now general consensus that phase V should be taken
as the diastolic pressure except when absent. 27,64,99
5.5 Cuffs
Modern cuffs consist of an inflatable cloth-enclosed bladder which encircles the arm and is
secured by Velcro or by tucking in the tapering end. The width of the bladder is
recommended to be about 40%, and its length 80%, of the arm circumference. Manufacturers
are now required to provide markings on the cuff indicating the arm circumference for which
it is appropriate (BS EN 1060-1) 21; these marks should be easily seen when the cuff is being
applied to an arm. When the bladder is too small (under-cuffing) it is possible to overestimate
blood pressure. The existence of over-cuffing and consequent underestimation is contentious
although likely to be of smaller magnitude.482,553,636
Conditions and environment
Blood pressure is maintained by a combination of mechanical, neuronal and endocrine selfregulating systems in the body. These systems can alter blood pressure in response to changes
in environment. Individual readings are influenced (for example) by age, ethnicity, disease,
the time of day, posture, emotions, exercise, meals, drugs, fullness of bladder, pain, shock,
dehydration, acute changes in temperature and changes in altitude. These influences can be
substantial, altering systolic readings by as much as 20 mmHg.65
41
Hypertension (partial update)
Measuring blood pressure
Standardising the environment in which blood pressure measurements are made reduces
variation and enhances the interpretation of a series of readings taken over time.27,99 A quiet,
comfortable location at normal room temperature is optimal. Ideally, the patient should not
need to pass urine, not recently have eaten, smoked or taken caffeine or exercise. Allowing
the patient to rest at least five minutes before measurement is also advised.27,65,99
Blood pressure readings tend to increase as patients move from the supine to standing
position. The change may not be significant, but it is traditional for measurements to be taken
whilst seated. Certain patients demonstrate a significant lowering of blood pressure when
standing (postural hypotension).27,65,66,99,452
Blood pressure readings also tend to increase as the patient's arm is lowered below the
horizontal and decrease when the arm is raised. When blood pressure is measured in the clinic
setting, the patient’s arm should be out-stretched, level with their heart and in line with their
mid sternum, and supported by a table or some other means.27,65,66,99,452 Blood pressure is
usually measured in the non-dominant arm, especially when using home or ambulatory
monitoring. Differences in readings may occur between arms. A BP difference of <10mmHg
can be considered normal, however, a difference of more than 20mmHg between arms is
unusual, occurring in <4% of people and is usually associated with underlying vascular
disease. Clinicians are advised to take readings in both of the patient's arms initially, and use
the arm with the higher reading for subsequent measurements of blood pressure. . Consistent
inter-arm differences of over 20/10 mmHg may suggest pathology warranting specialist
referral.27,65,99
5.6 White Coat Hypertension
42
Update 2011
The observation that clinicians (signified by their white coats) can cause spuriously high
blood pressure readings in patients was first described in the 1940s.58 Additionally,
sympathetic symptoms such as sweating, tachycardia and palpitation sometimes occur. The
effect is short-lived with blood pressure dropping to normality after or near the end of the
consultation. Consequently, a patient may present as hypertensive in clinic (in a primary or
secondary care setting) but be normotensive otherwise.
White Coat Hypertension (WCH) is reported to occur in as many as 15% to 30% of the
population,448 although this may be inflated due to inadequate evaluation of patients. It is
more common in pregnancy and with increasing age although poorly understood otherwise.569
The size of white coat effect in individuals can vary over time and a small proportion (4%)
may demonstrate atypical very high clinic readings.27 Failing to identify WCH makes
inappropriate treatment for hypertension in normotensive patients a possibility. Similarly,
hypertensive individuals can also exhibit WCH and may receive inappropriate dose titrations
or additional antihypertensive agents.490,506,635 Patients have historically been enrolled in trials
using clinic BP values, and these trials will almost certainly have included a proportion of
patients with WCH. It is unknown whether benefits of treatment differ substantially in those
with or without WCH.
“White Coat” Hypertension: A difference between clinic BP and home or ambulatory blood
pressure averages is expected. This difference has been reported to average approximately
10/5mmHg but this will vary considerably and is usually greater in people with a higher
baseline blood pressure and as people age. White coat hypertension is defined when a patient
has a persistently elevated clinic BP and a normal home or ambulatory BP day time average,
i.e. <135/85mmHg.
“White coat Effect” in people with hypertension: People with true hypertension, treated or
untreated, can also exhibit a “White Coat Effect”, for example a clinic BP reading that is
disproportionately greater than their home or ambulatory BP averages, but their home or
ambulatory BP averages are in a hypertensive range. Such patients are at risk of receiving
more BP medication than they need and will require out of office measurement to monitor the
efficacy of their BP treatment.
Hypertension (partial update)
Measuring blood pressure
5.7 Blood pressure measurement devices
5.7.1
5.7.3
Update 2011
5.7.2
There is considerable guidance about the range of appropriate devices for measuring blood
pressure.100,171,446 and about their maintenance and periodic recalibration [172 Local medical
physics and biomedical/clinical engineering departments can often give further advice.
Mercury sphygmomanometer
The mercury sphygmomanometer has been used for the traditional measurement of blood
pressure. It is reliable and provides the reference standard for indirect measurement. However
it is bulky, fragile and there are particular safety and economic concerns about the toxic
effects of mercury. Mercury is being phased out of clinical use and mercury
sphygmomanometers have already been removed from clinical areas in hospitals and primary
care. Thus, alternatives to mercury sphygmomanometry are now required for routine clinical
use.
Non-mercury devices that operate in a similar way to the traditional mercury column devices
are available and provide a suitable alternative to mercury devices when manual auscultation
is required to measure blood pressure.
Aneroid sphygmomanometers
Aneroid sphygmomanometers measure pressure using a lever and bellows system. They may
be less accurate than mercury sphygmomanometers and their alternatives (see above),
especially over time. Using the manual auscultation technique they are subject to the same
sources of observer error.64
Automated devices
Automated devices are increasingly being used in hospitals and primary care. All
sphygmomanometers need regular maintenance. Rubber tubing can crack and leak making
cuff deflation hard to control, underestimating systolic and overestimating diastolic readings.
Faulty valves can cause similar problems.64
Ambulatory blood pressure monitors
Ambulatory Blood Pressure monitoring (ABPM) involves a cuff and bladder connected to
electronic sensors which detect changes in cuff pressure and allow blood pressure to be
measured oscillometrically. The cuff is inflated by a battery powered compressor and sensors
within the cuff detect changes in pressure oscillations during cuff deflation. Systolic and
diastolic pressure readings are deduced from the shape of these oscillometric pressure changes
using an algorithm built into the measuring device. Developed as a research tool in the 1960s,
these devices have considerably reduced in size and now can be described properly as
ambulatory. Thus a patient's blood pressure can be automatically measured at repeated
intervals (commonly every 30 minutes) throughout the day and night, while they continue
routine activities. Systolic and diastolic pressure can be plotted over time, with most devices
providing average day, night and 24 hour pressures.448 (see Figure 2, page 41) An advantage
of ABPM is the removal of observer error with automated reading. However, oscillometric
measurement may be difficult in the presence of arrhythmias, particularly rapid atrial
fibrillation, and in a subgroup of the general population in whom oscillometric readings are
inaccurate for unknown reasons.445,448
A number of ABPM devices are available varying in size, weight, noise level, data
manipulation and cost.450,452 Devices should be independently validated to one or both of two
internationally accepted standards from the British Hypertension Society and the Association
for the Advancement of Medical Instrumentation.41,447,451 See British Hypertension Society
website www.bhsoc.org for a list of validated monitors.
When using ABPM, patients need some understanding of how the device works and
instruction about manual deflation, missed readings, arm position, and machine location:
fitting takes 15–30 minutes. An appropriately sized cuff is necessary as with non-ambulatory
Update
43
Hypertension (partial update)
Measuring blood pressure
monitoring and if one arm gives a higher reading at baseline then this should be used
subsequently. Patients may be asked to make diary records of events that are known to affect
blood pressure so that readings can be related to them, for example, periods of sleep. Sleeping
times can be recorded or fixed times may be predefined, including preparing for sleep (e.g.
9pm – midnight) and waking up (e.g. 6am – 9 am).448,450
Home blood pressure monitors
Home monitoring devices are oscillometric, measuring BP on the upper arm, the wrist or the
finger. Home monitoring potentially offers some similar benefits to ABPM. Frequent
measurement produces average values that may be more reproducible and reliable that
traditional clinic measurement. Potentially, white coat hypertension, systematic error,
terminal digit preference and observer prejudice can be removed.104,449,556 Home monitoring
allows patients to assess their own response to antihypertensive medication, which may
increase compliance with treatment. It has been argued that better evaluation provided by
home monitoring may reduce unnecessary treatment, increase compliance and thus deliver
cost savings.490,556 Home blood pressure devices are thought by some professionals to cause
anxiety or obsessive self interest.449,452,556,569
Potential disadvantages stem from the need for appropriate training to avoid biased
measurement. Use of inappropriately sized cuffs, isometric exercise when not resting the arm,
measurement after or during exercise and observer prejudice (for non-automated recording)
are possible.27 One study found that only 30% of patients using a manual home blood pressure
monitor correctly adhered to the protocol. Further, less than 70% of the self-reported
measurements were identical to those simultaneously recorded by the machine.303 Observer
bias was more apparent in those patients who were more hypertensive or whose readings
showed more variation. As with ABPM, home monitoring devices are oscillometric and may
have difficulty measuring pressure in cases of arrhythmias, and in certain patients for no
apparent reason.
See British Hypertension Society website www.bhsoc.org for a list of validated monitors.
Recommendations
1. Healthcare professionals taking blood pressure measurements need adequate initial training
and periodic review of their performance. [2004]
2. Because automated devices may not measure blood pressure accurately if there is pulse
irregularity (for example, due to atrial fibrillation), palpate the radial or brachial pulse
before measuring blood pressure. If pulse irregularity is present, measure blood pressure
manually using direct auscultation over the brachial artery. [new 2011]
4. When measuring blood pressure in the clinic or in the home, standardise the environment
and provide a relaxed, temperate setting, with the person quiet and seated, and their arm
outstretched and supported. [new 2011]
5. If using an automated blood pressure monitoring device, ensure that the device is
validateda and an appropriate cuff size for the person’s arm is used. [new 2011]
a
A list of validated blood pressure monitoring devices is available on the British Hypertension Society’s website (see
www.bhsoc.org). The British Hypertension Society is an independent reviewer of published work. This does not imply an
endorsement by NICE.
44
Update 2011
3. Healthcare providers must ensure that devices for measuring blood pressure are properly
validated, maintained and regularly recalibrated according to manufacturers’ instructions.
[2004]
Hypertension (partial update)
Diagnosis of Hypertension
6. In people with symptoms of postural hypotension (falls or postural dizziness):
 measure blood pressure with the person either supine or seated.
 measure blood pressure again with the person standing for at least a minute prior to
measurement. [2004, amended 2011]
7. If the systolic blood pressure falls by 20 mmHg or more when the person is standing:
 review medication
 measure subsequent blood pressures with the person standing
 consider referral to specialist care if symptoms of postural hypotension persist. [2004,
amended 2011]
Research recommendation
1. Which automated blood pressure monitors are suitable for people with hypertension and
atrial fibrillation?
Atrial fibrillation is common in older people and may prevent accurate blood pressure
measurement with automated devices. It would be valuable to know if this can be overcome.
6
Diagnosis of Hypertension
6.1 Predicting outcome using clinic, home and ambulatory measurements
6.1.1
Review question: In adults with suspected primary hypertension, what is the best method to
measure blood pressure (HBPM versus ABPM versus CBPM) to predict the development of
cardiovascular events?
Clinical evidence 2004
If clinic blood pressure measurements are inaccurate this may weaken the relationship
between blood pressure and cardiovascular risk. Studies were systematically identified and
45
Update 2011
Hypertension is diagnosed and subsequently treated to reduce the risk of developing stroke,
ischaemic heart disease, heart failure, peripheral vascular disease, renal disease, dementia and
premature death. A person’s risk is not only determined by their blood pressure but also by
the presence of target organ damage, established cardiovascular disease and other risk factors
for cardiovascular disease such as lifestyle (e.g. diet, smoking, obesity and lack of exercise),
diabetes and dyslipidaemia . The assessment of a person when contemplating a clinical
diagnosis of hypertension must take account of these additional factors which are discussed in
Chapter 7 of the guideline.
Blood pressure is highly variable and the 2004 guidance emphasised that hypertension should
not be diagnosed nor treatment offered on the basis of a single BP measurement.
Consequently, people with suspected hypertension have been required to undergo repeated
measurements of their clinic BP on repeated clinic visits to confirm or refute the diagnosis of
hypertension. The exception being the rarer occasions when patients present with severe
elevations of BP, usually associated with evidence of target organ damage, when treatment is
needed more urgently.
The emergence of automated BP monitoring, either for home use, or ambulatory BP
monitoring devices, has revealed that there can be marked discrepancies between clinic BP
measurement and home or ambulatory BP averages , which are known as either white coat
hypertension (see 5.6) or masked hypertension (where clinic BP is normal but ABPM and/or
HBPM measurements are elevated). The identification of these discrepancies has prompted
consideration as to whether the conventional clinic blood pressure measurement method is
still the most accurate at predicting the risk of future cardiovascular disease and establishing
the diagnosis of hypertension.
Hypertension (partial update)
Diagnosis of Hypertension
6.1.2
46
Update 2011
retrieved that prospectively compared the ability of ambulatory, home and clinic measures of
blood pressure to predict fatal or non-fatal cardiovascular events. Studies addressing markers
of evolving disease, such as left ventricular mass or hypertrophy, were not included because
of their uncertain relationship with patient outcome.
Details of six reports relating to four cohorts of patients were abstracted. Studies were
conducted in London, England,324 Ohasama, Japan,465,523 Umbria, Italy,526,613-615 and the final
cohort was provided by European patients enrolled in a drug trial.557 Two further studies are
ongoing.87,385,472
The four cohorts included about 4,500 participants; approximately 50% of participants were
male and their mean age was nearly 55 years. Most participants were Caucasian or Japanese
reflecting the location of the studies. The mean length of follow-up was five years.
The British study investigated ambulatory blood pressure using an intra-arterial cannula, and
thus its findings may not generalise to indirect ambulatory measurement. This limitation
accepted, 24 hour, day or night direct measurements predicted cardiovascular events whereas
clinic measurement did not.
The Ohasama study compared self-measured home BP and clinic BP. Neither method
demonstrated superior prediction of first stroke, although home measurement appeared to be a
better predictor of cardiovascular mortality.
In the Italian cohort, ambulatory 24-hour systolic blood pressure was a better predictor than
clinic assessment for cardiovascular morbidity and mortality. The analysis suggested that
white coat hypertension and nocturnal dipping are independently associated with the risk of
cardiovascular disease, the implication being that those not demonstrating a white coat effect
or nocturnal dipping are at greater risk. It is plausible that a nocturnal reduction in blood
pressure may protect target organs, although the definition of 'non-dippers' currently varies
between studies (examples include a mean nocturnal pressure fall of less than 10% or an
absolute reduction of less than 10/5 mmHg). Varying definitions, as well as classification of
day and night periods, may explain differences in the prevalence of non dippers seen in
studies.
The SYST-EUR trial enrolled 4,695 patients into a trial comparing calcium-channel blocker
initiated blood pressure control and placebo. A sub-study conducted in 46 of the 198
participating centres compared the prognostic value of ambulatory and clinic blood pressure
readings. When treatment and placebo groups were taken together, this study provided no
evidence that ambulatory values more accurately predicted cardiovascular morbidity or
mortality than clinic readings.
Combining the evidence from these four cohorts, the difference in prognostic accuracy of
home, ambulatory and clinic measures appears small and inconsistent. None of these studies
adequately described their approach to analysing their data or the statistical robustness of
models produced. A further potential confounder was the adequacy of clinic baseline
measurements. It is possible that SYST-EUR, which had better baseline clinic assessment,
minimised the 'regression to the mean' phenomenon and obtained more representative values.
On the other hand, it is clear from large epidemiological studies that there is a very precise
relationship between periodic clinic based blood pressure measurements and risk of
cardiovascular disease.361,379
Clinical evidence 2011
Three pooled analyses of prognostic studies210,254,326 and 11 individual prognostic
studies77,86,159,178,211,253,284,404,438,564 were found that fulfilled the inclusion criteria and looked at
the ability of clinic, home or ambulatory blood pressure measurements to predict outcomes.
Outcomes of interest were mortality, stroke, MI, heart failure, diabetes, vascular procedures,
hospitalisation for angina, and other major adverse cardiac and cerebrovascular events
(MAACE).
Hypertension (partial update)
Diagnosis of Hypertension
47
Update 2011
6.1.3
The three pooled analyses210,254,326 were meta-analyses of individual data from prospective
studies. The individual studies included in these pooled analyses were excluded from our
review in order to avoid duplication / double counting of data. Two of the pooled
analyses254,326 used data from four studies of random populations with longitudinal follow-up
of fatal and non-fatal CV outcomes. They both included the same studies, however the people
they included in the final analyses were different (one study326 excluded people with no nighttime data available, and the other study254 excluded people with no daytime data available).
The third pooled analysis210 used data from three studies in the Belgian Ambulatory Blood
Pressure Monitoring database (which contains individual data of HT patients from studies
performed in Europe and coordinated by the university of Ghent or Leuven). Patients had a
history of CV disease.
All prognostic studies were observational and were found to be methodologically sound /
have a low risk of bias (see quality assessment summary tables in appendix F). Studies that
were published before 2003 (the cut-off date of the original guideline, CG18436) were
excluded.
 Studies were categorised into those which compared:
 Home versus clinic measurements (five studies)86,211,438,534,564
 ABPM versus clinic measurements (11 studies)77,159,178,210,253,254,284,326,404
 ABPM versus home versus clinic measurements (two studies)211,534
Four studies were conducted in people who were known or suspected to have
hypertension86,159,178,404 and the rest of the studies were in population samples which would
have contained both hypertensive and non-hypertensive people. Mixed population studies are
a better representation of how BP monitoring would be used in clinical practice and the
prognostic ability of the blood pressure measurement methods to determine clinical outcome.
NOTE: The Hansen 2007 study254 only assessesd daytime ABPM measurements; the Dawes
2006 study159 only assessed 24h ABPM measurements; and the Fagard 2005 and Fagard 2008
studies210,211 only assessed daytime and night-time ABPM, and not 24h measurements. All
other studies assessed and compared separately all three types of ABPM measurements - 24h,
daytime and night-time). The protocol used for measuring blood pressure (for example, the
intervals between each ABPM reading and definitions of daytime and night-time periods)
varied between studies.
Evidence statements – clinical
The table below (Table 13) summarises the overall results of the prognostic studies included
for this review. Table 14summarises the numerical results for selected outcomes of the
prognostic studies included for this review. The full data for all outcomes can be found in the
evidence tables in the appendix.
NOTE: The ‘best method’ was chosen as the method of measuring BP that best predicted (ie.
statistically significant predictors and higher HR values) clinical outcomes (after adjustment
for covariates in multivariate analyses).
Table 13: Summary of included prognostic studies
Outcome
Best method
Representative of ‘real life’
home BP measurements?
4939
Mean 3.2 years
CV events
Home
Yes – measurements over 4 days
2081
Mean 6.8 years
Mortality and CV events
Home
Yes – measurements over 7
days; but home BP threshold
(for HT diagnosis) not given
665
Mean 8.2 years
CV events
NS difference
Yes – measurements over 3
days; but small study , and home
BP threshold (for HT diagnosis)
not given
Bjorklund 200477
872
Mean 6.6 years
CV morbidity
SBP: Office and
ABPM (daytime SBP
added more)
n/a
Dawes 2006159
10,129
Median 10 years
Mortality
ABPM (daytime)
n/a
5292
Mean 7.9 years
CV mortality
ABPM (especially
night-time)
n/a
Fagard 2008*210
302
Median 6.8 years
Mortality, CV mortality, CV events
ABPM (especially
night-time)
n/a
Hansen 2005253
1700
Up to 9.5 years
Mortality and CV mortality
ABPM
n/a
Hansen 2007*
7030
Median 9.5 years
CV death, stroke, cardiac events
and CHD
ABPM (CV events);
but no difference for
mortality (total and
CV)
n/a
Ingelsson 2006284
951
Up to 9.1 years
CHF
ABPM (night-time
DBP)
n/a
Kikuya 2007*326
5682
Median 9.5 years
CV death, stroke, cardiac events
and CHD
No difference
n/a
Mesquita-Bastos 2010404
1200
Mean 8.2 years
CV events and stroke
ABPM (especially
night-time)
n/a
Home vs clinic
Bobrie 200486
Niiranen 2010
438
Stergiou 2007564
48
Dolan 2005
178
254
Update 2011
ABPM vs clinic
Hypertension (partial update)
Follow-up time
Diagnosis of Hypertension
N
Study
Update 2011
Study
Representative of ‘real life’
home BP measurements?
Outcome
Best method
Fagard 2005211
391
Median 10.9 years
Major CV events
Home equal to ABPM
and better than office
No – home BP measurement
performed y investigator rather
than patient.
USega 2005534
2051
Mean 10.9 years
Mortality
No difference
No – only measured home BP
on 1 day; home BP threshold
(for HT diagnosis) not given
Home vs ABPM vs clinic
Hypertension (partial update)
Follow-up time
Diagnosis of Hypertension
N
49
Hypertension (partial update)
Diagnosis of Hypertension
Table 14: Summary of numerical results for prognostic studies (selected outcomes)
Outcome
Best
method
Bobrie 200486
CV events
Home
Home: 1.02 (1.01, 1.02) p=<0.001
Clinic: 1.01 (1.00, 1.01) p=0.09
Per 1mmHg rise in SBP
Niiranen 2010438
CV events
Home
Home: 1.22 (1.09, 1.37) p<0.001
Clinic: 1.01 (0.92, 1.12) p=0.80
per 10mmHg rise in SBP
Stergiou 2007564
CV events
No
difference
Home: 1.00 (0.99, 1.02) p=0.68
Clinic: 1.01 (0.99, 1.03) p=0.08
Per 1mmHg rise in SBP
Bjorklund
200477
CV
morbidity
SBP: Office
and ABPM
(daytime
SBP added
more)
ABPM (24h): 1.23 (1.07, 1.42) p<0.05
ABPM (daytime): 1.23 (1.07, 1.42) p<0.05
Clinic: 1.21 (1.04, 1.41) p<0.05
per 1SD rise in SBP
Dawes 2006159
Mortality
ABPM
(daytime)
ABPM (daytime): 1.51 (1.25, 1.83); p<0.001
Clinic: 1.02 (0.84, 1.24); p=0.90
highest quartile of SBP compared to ?lowest
Dolan 2005178
CV
mortality
ABPM
(especially
night-time)
ABPM (24h): 1.19 (1.14, 1.26) p<0.001
ABPM (night-time): 1.21 (1.16, 1.27) p<0.001
Clinic: 1.06 (1.02, 1.10) p<0.01
per 10mmHg rise in SBP
Fagard 2008*210
CV events
ABPM
(especially
night-time)
ABPM (24h): 1.20 (0.91-1.58) NS
ABPM (daytime): 1.03 (0.77-1.36) NS
ABPM (night-time): 1.34 (1.06-1.69) p<0.01
Per 1SD rise in SBP
Hansen 2005253
CV
mortality
ABPM
ABPM (24h): 1.51 (1.28, 1.77) p<0.0001
ABPM (daytime):1.50 (1.27, 1.76) p<0.0001
Clinic: 1.25 (1.10, 1.42) p<0.001
per 10mmHg rise in SBP
Hansen 2007*254
Cardiac
events /
CV events
ABPM (CV
events); but
no
difference
for mortality
(total and
CV)
Cardiac events ABPM (daytime): 1.13 (1.04, 1.23)
p<0.0001
Cardiac events Clinic: 1.06 (0.99, 1.13) p>0.05
CV events ABPM (daytime): 1.17 (1.10, 1.24) p<0.0001
CV events Clinic: 1.05 (1.00, 1.10) p>0.05
per 10mmHg rise in SBP
Ingelsson
2006284
CHF
ABPM
(night-time)
ABPM (24h): 1.13 (0.91, 1.40) p>0.05
ABPM (night-time): 1.21 (0.98, 1.49) p>0.05
Clinic: 1.25 (0.98, 1.59) p>0.05
per 1SD rise in SBP
Kikuya 2007*326
Cardiac
events
No
difference
ABPM (24hrs): 1.20 (1.13, 1.27) p<0.0001
ABPM (daytime): 1.16 (1.09, 1.23) p<0.0001
Clinic: 1.09 (1.04, 1.15) p<0.001
per 10mmHg rise in SBP
Mesquita-Bastos
2007404
CV events
ABPM (esp.
night-time)
ABPM (24h): 1.41 (1.20-1.65) <0.001
ABPM (daytime): 1.33 (1.10-1.60) <0.01
Study
HR (95% CI) for SBP measurement
Home vs clinic
ABPM vs clinic
Update 2011
50
Hypertension (partial update)
Diagnosis of Hypertension
Study
Outcome
Best
method
HR (95% CI) for SBP measurement
ABPM (night-time): 1.57 (1.32-1.86) p<0.001
Per 1SD rise in SBP
Home vs ABPM vs clinic
Major CV
events
Home equal
to ABPM
and better
than office
Home: 1.32 (1.06, 1.64) p=0.01
ABPM (daytime): 1.33 (1.07, 1.64) p<0.01
ABPM (night-time): 1.42 (1.16, 1.74) p<0.001
Clinic: 1.13 (0.88, 1.45) p=0.34
Per 1mmHg rise in SBP
Sega 2005534
Mortality
No
difference
No HRs given, but all entry BP values had a direct
exponential relationship with the risk of all-cause death or
CV death
Goodness of fit of the relationship of BP to risk of death
(CV and all-cause) was not less for clinic, compared to
home and ambulatory.
β Coefficient
ABPM (24h): 0.0557 ± 0.0008 p<0.0001
ABPM (daytime): 0.0479 ± 0.008 p<0.0001
ABPM (night-time): 0.0559 ± 0.007 p<0.0001
β Coefficient – the increase in risk per 1mm Hg increase in
SBP
Update 2011
Fagard 2005211
Summary
Studies showed that for predicting clinical outcomes:
ABPM versus CBPM (nine studies):
 ABPM was superior to CBPM (eight studies)
 There was no difference between ABPM and CBPM (one study)
HBPM versus CBPM (three studies):
 HBPM was superior to CBPM (two studies)
 There was no difference between HBPM and CBPM (one study)
HBPM versus ABPM versus CBPM (two studies):
 HBPM was similar to ABPM and both were superior to CBPM (one study)
 There was no difference between HBPM, ABPM and CBPM (one study)
6.2 Sensitivity and specificity of clinic, home and ambulatory measurements
6.2.1
51
Update 2011
Review question: In adults with suspected primary hypertension, what is the best method to
measure blood pressure (HBPM versus ABPM versus CBPM) to establish the diagnosis of
hypertension?
Clinical evidence
One systematic review/meta-analysis275 was found that fulfilled the inclusion criteria and
looked at the best method of measuring blood pressure for diagnosing hypertension. Studies
were included in the SR/MA if they were: RCTs, adult population (all ages), all settings
except hospitalised (the main focus was to be on primary care). Studies were excluded from
the SR/MA (unless these groups could be excluded from other data within a paper) if they:
did not specify the diagnostic thresholds used, had spectrum bias (no normotensives or
hypertensives in one measurement group), patients were pregnant, hospitalised, or were
receiving treatment at the time of the comparison. The systematic review/meta-analysis
included 20 studies (N=5863) and compared the sensitivity and specificity of CBPM and
HBPM measurements (using ABPM as the reference standard – as ABPM has been shown to
be the best blood pressure method for indicating prognosis). The systematic review/meta-
Hypertension (partial update)
Diagnosis of Hypertension
52
Update 2011
6.2.2
analysis was of good quality, however the quality of the studies it included ranged from poor
to good.
The population included in the 20 studies consisted of:
 primary care
 primary care at risk
 secondary care
 the general population
 general population at risk
 community volunteers
The 20 studies included in the SR/MA differed in terms of:
 Mean age (range <33 to 60 years)
 Gender: % male (range 16 to 69%)
 Sample size (range N=16 to N=2370)
 Mean baseline BP of population
 Sensitivity (Home vs ABPM range 0.48 to 0.91; clinic vs ABPM range 0.17 to 1.0)
 Specificity (Home vs ABPM range 0.34 to 0.92; clinic vs ABPM range 0 to 0.98)
 Number of measurements for ABPM (range: 24 to 111 in the daytime)
 Number of measurements for clinic BP (range: 2 to 18)
 Number of measurements for home BP (range: 18 to 56)
 Period of ambulatory measurement (range: 6 to 24 hours)
 BP thresholds used (range: ABPM SBP 91-144 mmHg; clinic SBP 90 to 160 mmHg; home
SBP 127 to 140 mmHg))
Quality assessment (QUADAS criteria) of the included studies showed that they:
 had good reporting of attrition
 had good selection criteria of participants
 had reporting bias: all studies had lack of clarity of reporting
 avoided both partial and differential verification bias (i.e. all patients in the studies
received the same comparison measurement tests, regardless of initial results)
 used validated devices for all strands of monitoring: 11/20 studies
 limited evidence of blinding to previous BP results from monitoring assessors
NOTE: only 10 of the 20 studies were ultimately included in the meta-analysis of data. Only
studies with the same reference test threshold and same index test threshold were pooled and
included in the meta analysis. Eight studies used a 135/85 mmHg ABPM threshold and a
140/90 mmHg clinic BPM threshold to diagnose hypertension, whilst three studies used a
threshold of 135/85 mmHg for both ambulatory and home diagnosis. However, one of the
clinic comparison studies used the full 24 hour mean ABPM rather than mean daytime
readings and was therefore not comparable to the others and excluded from the analysis.
Evidence statements – clinical
One SR/MA275 found the following sensitivities and specificities for CBPM and HBPM when
using ABPM as the reference standard (Table 15):
Hypertension (partial update)
Diagnosis of Hypertension
Table 15: CBPM and HBPM for diagnosing Hypertension. The thresholds used in the
SR/MA for diagnosis were: ABPM (daytime) 135/85 mmHg; clinic BP 140/90
mmHg; home BP 135/85 mmHg.
Statistical
significance (pvalue)
Clinic / ABPM
(7 studies)219,461,540,566,567,602,603
Home / ABPM
(3 studies)62,167,567
Sensitivity,% %
(95% CI)
74.62 (60.72, 84.83)
85.65 (77.95, 90.97)
NS (p-value not
reported)
Specificity, % (95%
CI)
74.61 (47.88, 90.38)
62.44 (47.98, 74.98)
NS (p-value not
reported)
Parameter / BP test
 Clinic versus Home BP (Table 15):
o there was NS difference between the BP measurement methods for sensitivity or
specificity
In a sensitivity analysis for CBPM which included only studies with mean BPs close to or
above the diagnostic threshold (ie. a typical general practice screening population with no
normotensives):
 CBPM sensitivity increased to 85.6% (CI 81.0 to 89.2) and specificity decreased to 45.9
(CI 33.0 to 59.3).
o NOTE: The home BP studies already used a typical general practice screening
population with no control group of normotensives and so the values remained the
same.
 This made HBPM the same as CBPM for sensitivity but better for specificity
Clinic BP thresholds (140/90 mmHg vs 150/90 mmHg);Table 16:
 sensitivity decreased with increasing BP threshold, however, the change was NS.
 specificity increased with increasing BP threshold, however, the change was NS.
Home BP thresholds (135/85 mmHg vs 140/90 mmHg and 130/80 mmHg);Table 16:
 Sensitivity significantly decreased with increasing threshold
 Specificity significantly increased with increasing threshold
Summary:
 Home BP is a better measurement than clinic BP for diagnosing HT (in a typical general
practice screening population), but is not as good as ABPM.
 A higher BP threshold (for clinic BP) resulted in worse sensitivity and better specificity for
diagnosing HT (compared to the current standard threshold used for diagnosis: 140/90
mmHg), however the effect was NS.
 A higher BP threshold (for home BP) resulted in a significantly worse sensitivity and
significantly better specificity for diagnosing HT (compared to the current standard
threshold used for diagnosis: 135/85 mmHg)
 A lower BP threshold (for home BP) resulted in significantly better sensitivity and
significantly worse specificity for diagnosing HT (compared to the current standard
threshold used for diagnosis: 135/85 mmHg)
53
Update 2011
Hypertension (partial update)
Diagnosis of Hypertension
Table 16: CBPM and HBPM – sensitivity and specificity of different thresholds for
diagnosing Hypertension. The thresholds used in the SR/MA for diagnosis by
ABPM (daytime) was 135/85 mmHg.
Test threshold
(referm=nces
not provided in
SR/MA)
Sensitivity, %
(95% CI)
Relative sensitivity,
% (95% CI)
Specificity, %
(95% CI)
Relative
specificity, %
(95% CI)
Clinic BP thresholds
140/90 (n=7)
74.73 (61.73 to
84.43)
1.00 (reference)
74.75 (49.82 to
89.82)
1.00 (reference)
150/90 (n=1)
66.34 (28.28 to
90.79)
0.89 (0.51 to 1.55),
p=0.68
86.16 (24.80 to
99.16)
1.15 (0.71 to 1.88),
p=0.57
Home BP thresholds
140/90 (n=1)
52.56 (34.71 to
69.78)
0.63 (0.45 to 0.88),
p=0.01
80.32 (67.88 to
88.74)
1.42 (1.20 to 1.68),
p<.0001
135/85 (n=3)
83.15 (76.09 to
88.45)
1.00 (reference)
56.68 (46.42 to
66.40)
1.00 (reference)
130/80 (n=1)
91.75 (84.37 to
95.82)
1.10 (1.03 - 1.18),
p=0.01
41.35 (30.13 to
53.53)
0.73 (0.57 to 0.93),
p=0.01
6.3 Cost-effectiveness of clinic, home and ambulatory measurements
6.3.1
Table 17: ABPM versus CBPM (diagnosis) – economic study characteristics
Study
Applicability
Limitations
Other Comments
Krakoff
2006338
USA
Partially
applicable(a)
Potentially
serious(b)
 CBPM diagnosed population.
 CBPM vs CBPM+ABPM at diagnosis.
 Decision analytic model incorporating prevalence of
white coat hypertension, rate of conversion to true
hypertension and drop-out rate from treatment.
 5-year time horizon.
 Costs: ABPM (diagnosis and annual follow-up) and
hypertension treatment.
a) Does not incorporate all relevant comparators. Does not incorporate health effects (possibly conservative towards
ABPM).Some uncertainty about the applicability of USA costs. Discounting not applied.
b) Source of prevalence of white coat hypertension unclear but varied in sensitivity analysis (15-20%). Limited sensitivity
analysis.
Table 18: ABPM versus CBPM (diagnosis) – economic summary of findings (mean per
person)
Study
Incremental
cost (£)
Incremental
effects
54
ICER
Uncertainty
Update 2011
Economic evidence – literature review
An economic evaluation should ideally compare all relevant alternatives. No studies were
identified comparing all of clinic blood pressure monitoring (CBPM), ambulatory blood
pressure monitoring (ABPM) and home blood pressure monitoring (HBPM) at diagnosis.
One study (Krakoff 2006338) was identified that examined the cost effectiveness of ABPM
compared with CBPM in the diagnosis of hypertension. This is summarised in the ABPM
versus CBPM economic evidence profile below (Table 17, Table 18). A full evidence table is
also provided in Appendix G: Evidence tables – health economic studies (2011 update).
One study was identified that examined HPBM and CBPM in the diagnosis of hypertension
but was excluded as it was judged to have serious methodological limitations.225
Hypertension (partial update)
Diagnosis of Hypertension
Study
Krakoff 2006
USA
338
Incremental
cost (£)
-£80(a)
Incremental
effects
N/a
ICER
Uncertainty
N/a
-£28 to -£132(b)
a) Converted from 2005 US dollars.
b) Two way sensitivity analysis varying white coat hypertension rate 15%-20% and the annual conversion rate of white coat
hypertension to true hypertension 5%-20%.
6.3.2
6.3.2.1
Economic evidence - original economic analysis
The GDG considered the clinical evidence reviewed as part of the guideline update to suggest
that ambulatory blood pressure monitoring (ABPM) may be more accurate at diagnosing
patients with hypertension than clinic blood pressure monitoring (CBPM) or home blood
pressure monitoring (HBPM); however it is also the most expensive option in terms of
monitor costs. HBPM was found to be more specific than CBPM but was also associated with
additional monitor costs. The use of ambulatory or home monitoring instead of clinic
monitoring to confirm a diagnosis of hypertension was identified as the highest economic
priority by the GDG due to it being a significant change in practice that would require
considerable investment in new devices by primary care.
As described above, no cost-effectiveness analyses comparing all of ABPM, HBPM and
CBPM were identified from the published literature. A protocol for a cost-effectiveness
analysis in development was submitted, in response to the call for evidence in this area (see
Methods), by a UK research groupb who had also undertaken a systematic review and meta
analysis of the sensitivity and specificity of CBPM and HBPM compared to ABPM that was
included in the guideline as part of the clinical evidence review275. However, the costeffectiveness analysis would not be completed within the timeframe of the guideline update
and so a collaboration was agreed between the GDG and the research group.
Below is a summary of the analysis that was undertaken. For full details please see Appendix
J: Cost-effectiveness analysis).
Methods
The population entering the model comprised people suspected of having hypertension based
on a screening clinic blood pressure reading. This group therefore included both those that
were truly hypertensive (true positive following screening) and those that were not (false
positive following screening). The diagnosis process aimed to correctly confirm both true
b
Richard McManus, Professor of Primary Care Cardiovascular Research, University of Birmingham; Sue Jowett, Senior
Lecturer in Health Economics, University of Birmingham; Pelham Barton, Reader in Mathematical Modelling,
University of Birmingham; James Hodgkinson, Research Fellow, University of Birmingham; Jonathan Mant, Professor
of Primary Care Research, University of Cambridge; Una Martin, Reader in Clinical Pharmacology, University of
Birmingham; Carl Heneghan, Reader in Evidence-Based Medicine, University of Oxford; Richard Hobbs, Head of
Primary Care Clinical Sciences, University of Birmingham.
55
Update 2011
A cost-utility analysis was undertaken to look at different blood pressure monitoring methods
for confirming a diagnosis of hypertension. A Markov model was used to estimate lifetime
quality-adjusted life years (QALYs) and costs from a current UK NHS and personal social
services perspective. Both costs and QALYs were discounted at a rate of 3.5% per annum in
line with NICE methodological guidance427. Uncertainty was explored through probabilistic
analysis and extensive sensitivity analyses.
The population used for the analysis was people with suspected hypertension – those with a
screening clinic blood pressure measurement equal or above 140/90 mmHg. Analyses were
run for ten gender and age (40, 50, 60, 70, 75 years) stratified subgroups.
The comparators selected for the model were confirmation of diagnosis with:
 Clinic blood pressure monitoring (CBPM)
 Home blood pressure monitoring (HBPM)
 Ambulatory blood pressure monitoring (ABPM)
Hypertension (partial update)
Diagnosis of Hypertension
hypertensives (in order to reduce their cardiovascular risk via treatment) and true
normotensives (in order to reduce unnecessary treatment). The key differences between
diagnostic options were their ability to accurately diagnose both these groups. One of the key
inputs in the model was therefore the sensitivity and specificity of the different diagnostic
options and this was based on the meta analysis275 included as clinical evidence in the
guideline. In addition the comparators varied in terms of the time they took to confirm a
diagnosis (and so receive treatment and the benefits of treatment in terms of cardiovascular
risk reduction).
Key model assumptions (these are discussed in more detail in the full write-up in Appendix J:
Cost-effectiveness analysis – blood pressure monitoring for confirmation of diagnosis of
hypertension):
 People with hypertension have a higher risk of cardiovascular events than people without
hypertension.
 Once a diagnosis of hypertension has been made (correctly and incorrectly; that is true
positives and false positives) people receive treatment including antihypertensive drugs.
 Only people who are truly hypertensive (true positives receive benefit in terms of
cardiovascular risk reduction from treatment.
o People who are truly normotensive but are treated (false positives) do not receive any
health benefits.
 People who are truly normotensive at entry to the model may develop hypertension over
time.
 People diagnosed as not hypertensive (correctly or incorrectly; that is true negatives and
false negative) will have a blood pressure check-up with CBPM every 5 years.
o At this check-up, it is assumed that they will again screen positive and so be suspected
of having hypertension again and their diagnosis is confirmed using the same method as
previously (CBPM, HBPM or ABPM)
 People who have had a cardiovascular event experience reduced quality of life and have an
increased risk of death.
Diagnosis confirmations using CBPM, HBPM or ABPM are associated with different initial
costs. As they also vary in terms of their ability to correctly diagnose people with and without
hypertension the downstream costs (including hypertension treatment, CVD costs and
checkups in those diagnosed as not hypertensive) and QALYs also vary.
Model inputs were based on the clinical effectiveness review undertaken for the guideline,
other published data and expert opinion where required. These are described in full in the
technical report in Appendix J. All model inputs and assumptions were validated by the GDG
and research group.
The cost of confirming a diagnosis with CBPM, HBPM and ABPM took into account device
costs, maintenance and healthcare professional time. In the base-case analysis the cost per
person was £38.00 for CBPM, £39.13 for HBPM and £53.40 for ABPM. This was based on
the following assumptions:
 CBPM was assumed to require at least a further two sets of readings should be taken at
monthly intervals. For costing purposes it was assumed in the base case that two sets of
readings would be taken; the first with a practice nurse and the second with a GP (as this
may involve a treatment consultation). A cost for the CBPM monitor was not included in
the costing as GPs will still require clinic monitors even if HBPM or ABPM at diagnosis in
instigated and so this cost will not vary dependant on the diagnosis strategy.
 HBPM was assumed to require measurements over 7 days. For costing purposes it was
assumed that two healthcare consultations would be required; an initial appointment with a
56
Hypertension (partial update)
Diagnosis of Hypertension
6.3.2.2
Results
This analysis of cost-effectiveness found that, confirming a diagnosis of hypertension with
ABPM instead of CBPM or HBPM was the most cost-effective option in all age/gender
subgroups (40, 50, 60, 70 and 75 years). In fact, ABPM was cost saving compared to CBPM
when long term costs were taken into account. The key driver of cost savings with ABPM
compared to CBPM was hypertension treatment costs avoided due to more accurate diagnosis
(increased specificity). Results are summarised in Table 19.
In most subgroups ABPM was associated with higher QALYs, as well as lower costs, than
CBPM and HBPM (that is ABPM was the dominant option). The exception was in the
subgroups with starting age 40 years and the female subgroup with staring age 50 years,
where ABPM still had lower costs but was associated with a small reduction in QALYs;
however, ABPM was still the most cost effective option in these scenarios.
Table 19: Basecase analysis results (probabilistic analysis) – cost effectiveness
(incremental costs and QALYS, and optimal strategy)
Incremental QALYs vs CBPM
Subgroup
Male, 40 years
Male, 50 years
Male, 60 years
Male, 70 years
Male, 75 years
Female, 40 years
Female, 50 years
Female, 60 years
Female, 70 years
Female, 75 years
HBPM
ABPM
Most CE
strategy
Proba
bility
CE
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
ABPM
100%
Incremental costs vs CBPM
ABPM
HBPM
-0.001
-0.004
-£48
-£235
(CI: -0.006, 0.004)
(CI: -0.009, 0.005)
(CI: -£128, £17)
(CI: -£322, £117)
0.001
0.006
-£34
-£156
(CI: -0.009, 0.009)
(CI: -0.003, 0.017)
(CI: -£89, £11)
(CI: -£233, -£62)
0.003
0.017
-£26
-£112
(CI: -0.010, 0.015)
(CI: 0.006, 0.029)
(CI: -£70, £7)
(CI: -£178, -£43)
0.005
0.022
-£23
-£89
(CI: -0.009, 0.017)
(CI: 0.012, 0.035)
(CI: -£65, £7)
(CI: -£150, -£30)
0.004
0.021
-£16
-£56
(CI: -0.007, 0.015)
(CI: 0.012, 0.030)
(CI: -£49, £6)
(CI: -£105, -£10)
-0.001
-0.006
-£68
-£323
(CI: -0.004, 0.001)
(CI: -0.008, -0.003)
(CI: -£167, £25)
(CI: -£389, £222)
-0.001
-0.001
-£40
-£182
(CI: -0.006, 0.004)
(CI: -0.006, 0.007)
(CI: -£106, £15)
(CI: -£256, -£79)
0.001
0.006
-£32
-£146
(CI: -0.006, 0.008)
(CI: 0.000, 0.015)
(CI: -£83, £11)
(CI: -£220, -£55)
0.003
0.014
-£20
-£82
(CI: -0.005, 0.011)
(CI: 0.008, 0.021)
(CI: -£59, £8)
(CI: -£142, -£25)
0.002
0.010
-£17
-£63
(CI: -0.004, 0.007)
(CI: 0.006, 0.015)
(CI: -£52, £11)
(CI: -£121, -£8)
CE= cost effective at a £20,000 threshold; CI = 95% confidence interval; QALYs = quality-adjusted life years.
57
Update 2011
practice nurse to explain to the patient how to use the monitor and a second once the
monitoring was complete with a GP to review the results and provide treatment advice if
necessary.
 ABPM was assumed to take place over a single 24 hour period. For costing purposes it was
assumed that two healthcare consultations would be required: an initial appointment with a
practice nurse to fit the monitor and a second with a GP to review the results and provide
treatment advice if necessary. In addition time for a nurse to download the ABPM data was
factored in.
 HBPM and ABPM device costs per person were calculated based on median published
costs to the NHS and assuming a lifetime of 5 years, no resale value, a discount rate of
3.5% and uses per year per machine of 40 and 125 respectively.
Alternative diagnosis costs were used in a series of sensitivity analyses. This included
scenarios with lower uses per year per machine and ABPM via direct access at hospital.
Hypertension (partial update)
Diagnosis of Hypertension
6.3.2.3
Update 2011
The conclusion that ABPM is cost-effective compared to CBPM and HBPM was robust to a
wide range of sensitivity analyses including those varying the cost of ABPM. As might be
expected, the conclusion was sensitive to changes to the accuracy of diagnosis with each
method and in some scenarios HBPM became the most cost-effective option. The conclusion
was somewhat sensitive to the assumption that check-ups for those diagnosed without
hypertension are undertaken every 5 years; in the two lower age subgroups HBPM became
cost-effective when check-ups were done annually. The conclusion was also sensitive to the
assumption that people who were not hypertensive but were treated did not receive benefits
from treatment; when non-hypertensive people also received a risk reduction from treatment
CBPM became the most cost-effective option as there was now benefit to misdiagnosing
people.
Interpretation & limitations
This analysis suggests that ABPM is the most cost-effective method of confirming a diagnosis
of hypertension in a population suspected of having hypertension based a CBPM screening
measurement >140/90 mmHg, compared with further CBPM or HBPM. This conclusion was
consistent across a range of age/gender stratified subgroups. Uncertainties in the analysis
were explored through extensive sensitive analysis which in most cases did not change
conclusions. Where conclusions were impacted this was discussed by the GDG and it was felt
that these should not change the overall conclusion.
It was noted that the analysis is most probably conservative in terms of ABPM in a number of
places. For example, ABPM reduces treatment costs compared to CBPM and HBPM and the
cost of these used in the basecase analysis is most likely on low side as it is based on most
commonly used generic drug costs and a single clinic visit per year. In addition, the basecase
does not incorporate any negative quality of life impacts of being on treatment and when even
a 1% reduction in quality of life is incorporated into the analysis QALYs differences between
options are considerably more favourable for ABPM. These effects were omitted from the
basecase analysis because side effects of antihypertensive drugs are generally fairly mild and
rare and patients can often change drugs if they experience side effects but also because no
appropriate data was identified to quantify any effects. However, it is not implausible that
there may be a small negative impact of being on pharmacological treatment due to side
effects.
In was noted in GDG discussions that there were potentially some additional benefits of
ABPM that were not captured by the model but that would be valued by patients. With
ABPM less people are incorrectly diagnosed as having hypertension when they do not. These
patients will therefore avoid unnecessarily drug treatment which will mean they won’t
experience side effects, incur prescription costs or be labelled as having a medical condition,
with the potential psychological and practical impacts this can have305. With ABPM patients
will also get a definitive diagnosis more quickly that with CBPM.
58
Update 2011
Sensitivity and specificity inputs
The relative sensitivity and specificity of CBPM, HBPM and ABPM is the key differentiator
between treatments in the model and as such is an important input.
However, there were a number of limitations to the estimates of sensitivity and specificity
used in the model.
A key assumption in the model, and the meta analysis used for sensitivity and specificity
estimates, was that ABPM is the reference standard for diagnosing hypertension and so has
100% sensitivity and specificity. This is a potential limitation in that ABPM probably does
not have 100% sensitivity and specificity. However, prognostic studies indicated that ABPM
was most predictive of prognosis and so this was considered a reasonable assumption for the
analysis; without making this assumption it would not be possible to undertake the analysis.
Hypertension (partial update)
Diagnosis of Hypertension
Conclusions were however somewhat sensitive to variations in the sensitivity and specificity
values, with HBPM becoming cost effective in some scenarios. However, while there is
uncertainty around the assumption that ABPM is the gold standard with 100% sensitivity and
specificity, the instances when conclusions were changed were generally quite extreme. For
example, when the sensitivity and specificity of ABPM were set equal to that of HBPM or
when the sensitivity of HBPM was increased to 100%.
In addition, while it is known that sensitivity and specificity vary with disease prevalence
(and so age) data was not available to allow this to be incorporated into the basecase analysis.
However, when examined in exploratory sensitivity analyses it seemed that it would probably
not impact conclusions.
The GDG carefully considered the uncertainty around the estimates of sensitivity and
specificity but given the currently available evidence felt that it should not impact the overall
conclusion that ABPM was the preferred option.
Differential treatment initiation threshold
In the model it is assumed for practical reasons that all people diagnosed with hypertension
(CBPM 140/90 mmHg; HBPM/ABPM 135/85 mmHg) receive pharmacological treatment.
59
Update 2011
Treating those who are not hypertensive
The basecase conclusion that ABPM was a more cost-effective option for confirming a
diagnosis of hypertension than CBPM or HBPM was sensitive to the assumption that only
people who were hypertensive received benefits (cardiovascular risk reduction) from
treatment. When a risk reduction was also applied to people who were treated but who were
not hypertensive (people incorrectly diagnosed as having hypertension), CBPM was the most
cost effective option across all subgroups.
The basecase assumption was based on the clinical GDG members’ opinion that there is
currently insufficient evidence of benefit for initiating treatment below the currently
recommended thresholds. While there is evidence of a continuous relationship between blood
pressure and cardiovascular risk361, it is not well established that initiating blood pressure
treatment below 140/90 mmHg reduces that risk in people with uncomplicated hypertension.
The meta analysis reported by Law and colleagues351 was used to inform the cardiovascular
risk reduction in the model for people with and without hypertension as results were stratified
by pre-treatment blood pressure; people with hypertension therefore got a greater risk
reduction than people without in the analysis. This meta analysis was reviewed as part of the
guideline update in relation to the question of what the treatment initiation threshold should
be (Chapter 0). This analysis asserts that cardiovascular risk reduction is obtained at all levels
of pre-treatment blood pressure. However, the GDG noted that the analysis included studies
with a range of populations and those that provided information for risk reduction where pretreatment blood pressure was below 140/90 mmHg were generally in populations with a
history of cardiovascular disease or other increased risk that are not necessarily representative
of the more general hypertension population.
The sensitivity analysis results, with CBPM more cost-effective than ABPM or HBPM,
suggests that misdiagnosing people as having hypertension when they do not is a good thing
because the health benefits of doing so are worth the additional cost of treatment. This result
is therefore more to do with what the diagnostic threshold should be rather than the method
that should be used to confirm diagnosis. It should also be noted that potential negative effects
of treatment (in terms of reducing people quality of life) were not considered in this
sensitivity analysis.
The basecase analysis reflects the GDG’s interpretation of the clinical data relating to
treatment thresholds and as such was considered to reflect the most appropriate analysis for
informing which method should be used to confirm a diagnosis of hypertension.
Hypertension (partial update)
Diagnosis of Hypertension
However, this guideline recommends a differential treatment initiation threshold whereby
people diagnosed with hypertension (by the above definition) generally receive
pharmacological treatment if their blood pressure is >160/100 mmHg (HBPM/ABPM
>150/95 mmHg), or they have an estimated 10-year cardiovascular risk equivalent to 20% or
greater, target organ damage, pre-existing cardiovascular disease, renal disease or diabetes. In
those with hypertension but not eligible for pharmacological treatment it is recommended
they receive lifestyle advice and an annual check-up.
The implications of this simplification are likely to be that the analysis somewhat
overestimates the costs of treating hypertension as some people won’t need to be treated and
somewhat overestimates the benefits of treatment (QALY gain), as some people won’t get
treated and so won’t get the risk reduction from treatment. However, the cost implications
will be mitigated by the fact that many people will eventually need drug treatment and that
nearly half the cost of hypertension treatment in the model is the annual check-up which will
still be required in those that have hypertension but not receiving drug treatment. The
treatment costs used in the basecase analysis are also potentially conservative. In addition, the
QALYs implications will be mitigated by the fact that the people who do not receive
treatment will be at lower risk so the people who remain in the model will have higher risk
and benefit more on average and lifestyle advice will provide some risk reduction in some
patients at least.
In addition to the above considerations, the implication of the differential pharmacological
treatment initiation threshold is effectively a reduction in the number of people eligible for
treatment. This is therefore somewhat addressed by the sensitivity analysis where the
prevalence of true hypertension in the model is varied through a wide range. The conclusion
that ABPM was the most cost-effective option was maintained through a prevalence of true
hypertension is the suspected hypertension population of 10-80%.
6.3.3
Model input uncertainty
Throughout this report it has been highlighted where there have issues with model input
uncertainty – this is a limitation of the analysis. In some places there was a lack of data to
inform inputs; this included CVD event and post-event costs and the prevalence of true
hypertension in a population of people with suspected hypertension. In other places there was
variability between settings or patients, such as the cost of ABPM and the frequency of checkups in those diagnosed without hypertension. The best available or more likely inputs were
used for the basecase analysis and these were varied in sensitivity analyses.
Evidence statements – economic
 One partially applicable study with potentially serious limitations found that ABPM was
cost saving compared to CBPM; the treatment costs avoided from not treating patients with
WCH were greater than the additional costs of ABPM.
 New economic analysis from a current UK NHS and PSS perspective comparing CBPM,
HBPM and ABPM for confirming a diagnosis of hypertension in a population with
suspected hypertension found ABPM to be the most cost effective option across a range of
60
Update 2011
Check-up frequency
In the basecase analysis it was assumed that people who were diagnosed without hypertension
were checked-up every 5 years. In a sensitivity analysis where this was change to an annual
check-up, ABPM was no longer cost-effective in younger age groups. The GDG discussed
the implications of this finding and felt that, while check-up frequency will vary between
patients, on balance this should not impact the overall conclusion that ABPM should be used.
It was however noted that in younger patients diagnosed as not hypertensive but in whom
frequent follow-up is planned, it might be considered reasonable to use an alternative to
ABPM to avoid high diagnosis costs.
Hypertension (partial update)
Diagnosis of Hypertension
age subgroups in both men and women. In most subgroups ABPM was found to both
improve health (increased QALYs) and reduce costs overall. The conclusion was robust to
the majority of sensitivity analyses undertaken including those varying the cost of ABPM.
6.4 Measurement protocols for diagnosing hypertension
6.4.1
6.4.1.1
Ambulatory blood pressure measurement
Review question: In adults with primary hypertension, what protocol should be used when
measuring ambulatory blood pressure for treatment and diagnosis?
Clinical evidence
The literature was searched for all years (as this was not addressed in the previous
guidelines)425,436 and all study types were included. Studies were excluded if the population
consisted of people who were exclusively diabetic or had CKD. Validation studies of ABPM
machines were also excluded.
53 studies77,88,111,151,178,190,200,210,211,237,253,271,272,284,325,326,363,387,405,416,456,491,534,562,563,573,622
46,52,56,114,131,133,150,196,353,386,389,390,420,473,527,530,531,538,541,557,576,595,600,608,609,654
were found that
fulfilled the inclusion criteria and assessed what protocol should be used when measuring
ambulatory BP for the treatment and diagnosis of adults with primary hypertension.
The studies addressing the question were categorised into two different types:
1. Prognostic studies (17studies;17 papers)77,88,131,178,210,211,237,253,284,325,326,363,405,491,534,557,576 –
those that assess the prognostic significance of ambulatory BP and the optimal schedule for
measurement based on outcome data
2. Reliability / reproducibility studies (36 studies; 36
papers)46,52,56,111,114,133,150,151,190,196,200,271,272,353,386,387,389,390,416,420,456,473,527,530,531,538,541,562,563,573,59
5,600,608,609,622,654
- those that assessed any of the following - the optimal ambulatory BP
schedule based on:
a) the reproducibility of ABPM
b) its stability over time (variability of BP over time)
c) the relationship (correlation) between day and night values with mean 24h ABPM
values
d) its ability to identify people diagnosed with HT / NT / ICH or dippers and non-dippers
e) changes in BP in response to treatment
Reliability /repeatability studies were deemed to be applicable to the question because they
showed which aspects of the ABPM protocol (daytime, night-time, or 24h blood pressure
measurements) were the most reliable, and therefore served as an indication of the ‘best’ /
optimal ABP measurements to be taken.
Details of all the studies are included in Table 20and Table 26.
61
Hypertension (partial update)
Diagnosis of Hypertension
62
Update 2011
Table 21summarises the numerical results for selected outcomes of the prognostic studies
included for this review. The full data for all outcomes can be found in the evidence tables in
the appendix. A summary of the measurement intervals for BP readings used by each of the
studies is summarised in Table 20, Table 22 and Table 23. All prognostic studies were found
to be methodologically sound / have a low risk of bias (see quality assessment summary
tables in appendix F) except for the Li 2008 study363 which was rated as ‘unclear’ for a
number of potential methodological flaws.
NOTE: For the prognostic studies, the ‘best method’ was chosen as the method of measuring
BP that best predicted (ie. statistically significant predictors and higher HR values) clinical
outcomes (after adjustment for covariates in multivariate analyses). For the
‘reproducibilty/reliability studies’ the ‘best method’ was chosen as the the method / protocol
of measuring blood pressure that was the most reliable or repeatable.
Table 20: Study details and results for prognostic studies assessing the optimal ABPM protocol
Reference / study
type
Bjorklund et al.,
200477
CV mortality
Proposed protocol (authors’ conclusions)
– best prognostic ability
24h, daytime and night-time are all
predictors
Use SBP not DBP
D – range 15-30 mins
N – range 30-60 mins
Total mortality,
CV mortality,
non-CV mortality,
CV events, stroke,
cardiac events
Both daytime and night-time BP (need to
record ABPM throughout the whole day).
NOTE: 24h BP was not measured.
Median 5
years
D – 30 mins
N – <60 mins
Total mortality,
CV mortality, CV
events, MI, stroke
24h and daytime (are better than night-time,
especially SBP)
OSC
Mean 7.9
years
every 30 mins
All-cause
mortality; Cardiac
mortality; CV
mortality
Night-time (better than daytime or 24h)
General
population in
primary care
practice (HT
and NT)
-
Median 10.9
years
D – 15 mins
N – 30 mins
CV events
Night-time (better than daytime)
HT (with
history of CV
not
specified
Median 6.8
years
D –range 15-30 mins
(10am – 6pm)
Follow-up
time
Mean 6.6
years
Time and frequency
of measurement
every 20 mins
OSC or
AUS
Median 9.6
years
HT
-
5292
HT
391
302
N
Population
Device
872
General
population
(HT and NT)
AUS
7458
analy
sed
General
population
(HT and NT)
2232
Outcomes
within-group
comparison
Boggia et al., 200788
Clement et al., 2003
131
Within-group
comparison
Dolan et al., 2005178
within-group
comparison
Fagard et al.,
2005211
within-group
comparison
Update 2011
Fagard et al.,
2008210
Update 2011
63
Pooled analysis of
other study data,
within-group
comparisons
(IDACO)
Hypertension (partial update)
Diagnosis of Hypertension
Prognostic studies
AUS
Mean Mean
84 months
D – 15 mins
N – 15 or 30 mins
CV complications
24h, daytime, night-time and arising BP are
all predictors (24h, daytime and arising
slightly stronger predictors)
Single BP value on rising in the morning (is
as good as mean daytime or mean 24h
measurements)
Use SBP not DBP
General
population
(HT and NT)
OSC
Up to 9.5
years
D – 15 mins
N – 30 mins
All-cause
mortality; CV
mortality
Night, day and 24h SBPs and DBPs
DBP better than SBP
951
General
population
(HT and NT)
AUS
Up to
9.1years
(mean range
0.1 – 11.4
years)
D – 20 or 30 mins
N – 20 or 60 mins
CHF
Night-time (better than daytime or 24h)
688
HT
Intraarterial
ABPM
Mean 9.2
years
Every hour
Non-CV death,
coronary death,
CeV death,
peripheral
vascular death,
nonfatal MI,
nonfatal stroke,
coronary
revascularisation.
24h, daytime and night-time all predictors
SBP and DBP in age <60
Only SBP in age >60
5682
General
population
-
Median 9.5
years
1 study: every 20 mins
1 study: every 30 mins
CV events;
coronary events;
Device
256
HT
1700
Outcomes
All-cause
mortality; CV
mortality;
composite of
major CV events
within-group
comparison
Hansen et al.,
2005253
64
within-group
comparison
Ingelsson et al.,
2006284
within-group
comparison
Khattar et al.,
2001325
within-group
comparison
Kikuya et al.,
2007326
24h, daytime and night-time (SBP and DBP)
Hypertension (partial update)
Time and frequency
of measurement
N – range 30-60 mins
(12am – 6am)
Population
disease)
Pooled analysis of
other study data
,within-group
comparisons
Gosse et al., 2001237
Proposed protocol (authors’ conclusions)
– best prognostic ability
Night-time
Follow-up
time
N
Diagnosis of Hypertension
Reference / study
type
Population
(HT and NT);
<10% had
underlying
CV disease
Device
Follow-up
time
Time and frequency
of measurement
1 study: 15 mins day,
30 mins night
1 study: 20 mins day,
45 mins night
7458
General
population
(HT and NT)
not
specified
Median 9.6
years
D – interval not
specified
N – interval not
specified
General
population
(HT and NT)
OSC
Mean 10.6
years
30 mins over 24 hours
Pooled analysis of
other study data,
within-group
comparisons
(IDACO)
Li et al., 2008363
Summary of
prospective
population studies
(case series)
Metoki et al.,
2006405
1542
65
Weekday
within-group
comparison
Pickering et al.,
2007491
CV mortality,
non-CV mortality,
CV events, stroke,
cardiac events
Proposed protocol (authors’ conclusions)
– best prognostic ability
Daytime and night-time (depending on
which outcome)
Night-time better for mortality outcomes
Daytime better for non-CV mortality
Both for CV events and stroke
Need to record ABPM throughout the whole
day
Mortality risk
from CeV and CV
events
Night and early morning 2h SBP (CeV and
CV mortality)
Elevated daytime 2h SBP (Haem stroke
mortality)
elevated night-time 2h SBP (cerebral
infarction and HD mortality)
High BP at different times of day is
associated with different subtypes of CeV
and CV mortality risk.
average of 4 SBP = 2hr
SBP value at different
periods
8945
1 study:
general
population
(HT and NT)
6 studies: HT
(NT controls)
OSC or
AUS
Mean 5.8
years
15-30 mins over 24
hours
Cardiac events;
stroke
Daytime for cardiac events, night-time for
stroke
One summary measure not enough to predict
different clinical outcomes
2051
General
population
(HT and NT)
OSC
Mean 10.9
years
every 20 mins
All cause
mortality; CV
mortality
Nighttime better than daytime
SBP better than DBP
Summary of
prospective
population studies
(case series)
Sega et al., 2005534
Outcomes
cardiac events;
fatal/non-fatal
stroke
Hypertension (partial update)
N
Diagnosis of Hypertension
Reference / study
type
Time and frequency
of measurement
OSC
Mean 4.4
years
D - ≤ 30 mins
N - ≤ 30mins
Total mortality,
CV mortality, CV
events, stroke,
cardiac events
Night-time (better than daytime)
Excluding the first 2h does not improve
accuracy
OSC
Mean 51.5
months
D – 30 mins
N – 30 mins
CV events
Higher 24-h and nighttime BP (SBP and
DBP) are associated with a higher incidence
of CV events
Population
Device
837
HT (ISH)
324
HT and NT
Within-group
comparison:
substudy ofSyst-Eur
trial
Suzuki et al.,
2000576
Outcomes
Within-group
comparison
66
NT = normotensives; HT = hypertensives; ISH = isolated sytolic HT; AUS = auscultatory device; OSC = oscillometric device; D = daytime; N = night-time
Hypertension (partial update)
Staessen et al.,
1999557
Proposed protocol (authors’ conclusions)
– best prognostic ability
Follow-up
time
N
Diagnosis of Hypertension
Reference / study
type
within-group
comparison
(PAMELA study)
Bjorklund et al., 2004
77
Outcome
HR (95% CI) for SBP measurement
CV mortality
ABPM (24h): 1.23 (1.07, 1.42) p<0.05
ABPM (daytime): 1.23 (1.07, 1.42) p<0.05
ABPM (night-time): 1.18 (1.03, 1.34) p<0.05
per 1SD rise in SBP
Boggia et al., 200788*
CV events
ABPM (24h): not given
ABPM (daytime): 1.16 (1.07-1.26) p<0.001
ABPM (night-time): 1.21 (1.12-1.30) p<0.001
Per 1SD rise in SBP
67
Clement et al., 2003 131
CV events
No HRs given. Relative Risks:
ABPM (24h): 1.34 (1.11-1.62)
ABPM (daytime): 1.30 (1.08-1.58)
ABPM (night-time): 1.27 (1.07-1.51)
Per 1SD rise in SBP
Dolan et al., 2005178
CV mortality
ABPM (24h): 1.19 (1.14, 1.26) p<0.001
ABPM (daytime): 1.15 (1.10, 1.21) p<0.001
ABPM (night-time): 1.21 (1.16, 1.27) p<0.001
per 10mmHg rise in SBP
Fagard et al., 2005211
CV events
ABPM (24h): Not given
ABPM (daytime): 1.33 (1.07, 1.64) p<0.01
ABPM (night-time): 1.42 (1.16, 1.74) p<0.001
Per 1mmHg rise in SBP
Fagard et al., 2008210*
Composite of major
CV events
ABPM (24h): 1.20 (0.91-1.58) NS
ABPM (daytime): 1.03 (0.77-1.36) NS
ABPM (night-time): 1.34 (1.06-1.69) p<0.01
Per 1SD rise in SBP
Gosse et al., 2001237
CV complications
No HRs given,only characteristics of people with vs without complications and the statistical difference.
ABPM (24h): 133 ± 16 vs. 143 ± 14 (p<0.001)
ABPM (daytime): 138 ± 16 vs 149 ± 15 (p<0.01)
ABPM (night-time): 121 ± 17 vs 129 ± 14 (p<0.05)
SBP mm Hg without vs with complications Mean±SD
Hypertension (partial update)
Study
Diagnosis of Hypertension
Table 21: Summary of numerical results for prognostic studies (for selected outcomes)
HR (95% CI) for SBP measurement
CV mortality
ABPM (24h): 1.51 (1.28, 1.77) p<0.0001
ABPM (daytime):1.50 (1.27, 1.76) p<0.0001
ABPM (night-time): 1.41 (1.23, 1.62) p<0.0001
per 10mmHg rise in SBP
Ingelsson et al., 2006284
CHF
ABPM (24h): 1.13 (0.91, 1.40) p>0.05
ABPM (day-time): 1.08 (0.85, 1.36) p>0.05
ABPM (night-time): 1.21 (0.98, 1.49) p>0.05
per 1SD rise in SBP
Khattar et al., 2001325
all cause mortality. (no
results for cornonary
death)
<60 yrs ABPM (24h): 1.01 (1.00, 1.02)p=0.04
< 60 yrsABPM (daytime): 1.01 (1.00, 1.02)p=0.04
<60 yrs ABPM (night-time): 1.01 (1.00, 1.02)p=0.04
>60 yrs ABPM (24h): 1.02 (1.00, 1.03) p=0.003
>60 yrsABPM (daytime): 1.02 (1.00, 1.03)p=0.004
>60 yrs ABPM (night-time): 1.02 (1.00, 1.03) p=0.007
No info on the reference rise of SBP, but likely per 1mmHg
Kikuya et al., 2007326
CV events – defined as
CV endpoints in the
evidence table (also
used cardiac events in
red)
ABPM (24hrs): 1.24 (1.19, 1.30) p<0.0001
ABPM (daytime): 1.20 (1.15, 1.25) p<0.0001
ABPM (night-time): 1.18 (1.14, 1.23) p<0.0001
ABPM (24hrs): 1.20 (1.13, 1.27) p<0.0001
ABPM (daytime): 1.16 (1.09, 1.23) p<0.0001
ABPM (night-time): 1.16 (1.10, 1.22) p<0.0001
per 10mmHg rise in SBP
Hansen et al., 2005
68
Li et al., 2008363*
CV events
ABPM (24h): not given
ABPM (daytime): 1.16 (1.07-1.26) <0.001
ABPM (night-time): 1.21 (1.12-1.30) <0.0001
per 1SD rise in SBP
Metoki et al., 2006405
Mortality risk from
CeV and CV events
ABPM (24h): 1.76 (1.39-2.25) p<0.002
ABPM (daytime): 1.59 (1.25-2.01) p<0.002
ABPM (night-time): 1.78 (1.40-2.27)p<0.002
Per 1SD rise in SBP
Pickering et al., 2007491*
Cardiac events
ABPM (24h): not given
Hypertension (partial update)
Outcome
253
Diagnosis of Hypertension
Study
Sega et al., 2005534
CV mortality
No HRs given, but all entry BP values had a direct exponential relationship with the risk of all-cause death
or CV death
Goodness of fit of the relationship of BP to risk of death (CV and all-cause) was not less for clinic,
compared to home and ambulatory.
β Coefficients:
ABPM (24h): 0.0557 ± 0.0008 p<0.0001
ABPM (daytime): 0.0479 ± 0.008 p<0.0001
ABPM (night-time): 0.0559 ± 0.007 p<0.0001
β Coefficient – the increase in risk per 1mm Hg increase in SBP
Staessen et al., 1999557
CV events
ABPM (24h): 1.20 (0.98-1.49) NS
ABPM (daytime): 1.17 (0.96-1.44) NS
ABPM (night-time): 1.23 (1.03-1.46) ≤0.05
per 10mmHg rise in SBP
Suzuki et al., 2000576
CV events
ABPM (24h): 1.28 (1.05 to 1.54) p< 0.05
ABPM (daytime): No HR reported
ABPM (night-time): 1.34 (1.13 to 1.58)p < 0.01
per 10mmHg rise in SBP
Hypertension (partial update)
HR (95% CI) for SBP measurement
ABPM (daytime): HR = 1.29(95% CI: 1.20-1.39); p < 0.0001
ABPM (night-time): HR = 1.22(95% CI: 1.14-1.30); p < 0.0002
per 10mmHg rise in SBP
Diagnosis of Hypertension
Outcome
69
Study
Reference / study
type
Antivalle et al.,
199046
Frequency of measurements
N
Population
Device
Follow-up
22
HT
AUS
and
OSC
4 weeks
(3
measurement
s: baseline, 2
and 4 weeks)
case-series: RCT
substudy
Consec
utive
reading
s
24h
Time of measurement
Mathematical
method
Proposed number of
measurements (authors’
conclusions)
Daytime
Reproducibility
of BP (between
the 3
measurements
over time)
Differences in BP measurements (3
measurements) was only significant
during waking hours
Long-term
reproducibility
of BP (between
the 2
measurements
over time): SD
Daytime BPwas better (vs nighttime and 24h)
Reproducibility
of BP (between
the 2
measurements
over time, after
placebo
treatment)
Placebo administration resulted in
SS reductions between baseline and
1 month 24h ABPM (SBP), and
daytime SBP/DBP.
No treatment resulted in NS
differences between baseline and 1
month for 24h, daytime and nighttime SBP/ DBP.
This suggests a placebo effect on
BP.
ABPM for 48 h revealed a
Night-time
24h
70
intervals not given
Asagami et al.,
199652
64
Borderline
HT
AUS
and
OSC
within-group
comparison
Asmar et al.,
200156
24h
on a work
day
Daytime (30 mins)
Night-time (1 hr)
24h
30
HT
-
RCT
Calvo et al.,
1-2 years
823
HT
OSC
1 month
(2
measurement
s1 month
apart)
24h
48 h
48h
Daytime (15 mins)
Night-time (30 mins)
24h
D – 20 mins (07.00-
Comparison of
Hypertension (partial update)
Table 22: Study details and results for reliability/reproducibility studies assessing the optimal ABPM protocol
Diagnosis of Hypertension
Reliability and reproducibility studies
2003111
23.00)
Case-series
N – 30 mins (23.0007.00)
day-to-day
variations in BP
statistically significant pressor
response (this could largely be due
to the novelty of wearing an ABPM
device for the first time).
The pressor effect remains
statistically significant for the first
10 h of monitoring, independent of
gender, day of the week of
monitoring and number of a-HT
drugs used.
Nocturnal mean BP was similar
between both days of sampling.
The effect diminished, but was not
eliminated, in extent and duration
for successive sessions of
ambulatory monitoring.
ABPM for just 24 h may be
insufficient for a proper diagnosis
of HT, evaluation of treatment
efficacy and identification of
dipping status in relation to targetorgan damage.
Reproducibility
of BP (between
the 2
measurements
over time)
24h BP was more reproducible
over time than daytime and nighttime BP measurements.
Reproducibility
of BP (between
the 2
measurements
over time)
Average daytime ABPM DBP was
more reproducible than a single
measuremnt from daytime. There
was improved reproducibility with
more measurements during the day
ABPM started on a
weekday (Mon, Wed or
Fri)
71
Campbell et al.,
2010114
72
HT and NT
OSC
within-group
comparison
Coats et al., 1992
within-group
comparison
2 years (2
measurement
s 2 years
apart)
24h
Daytime (15 mins)
Night-time (30 mins)
24h
133
100
HT
-
1 month
(2
measurement
s1 month
apart)
24h
Daytime only (30 mins)
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
208
HT
OSC
case-series
Cuspidi et al.,
2007151
611
ICH
OSC
3 weeks
(2
measurement
swithin 3
weeks)
24h
2 x 24h
periods (1-4
weeks apart)
24h
Daytime (15 mins)
Night-time (20 mins)
24h
D (working day) – 15
mins (07.00-23.00)
N – 20 mins (23.0007.00)
Case-series
72
Eguchi et al.,
2010190
43
HT
OSC
Measuremen
ts twice
within a 2week interval
between
measurement
s
24h
Every 30 mins
80
HT and NT
OSC
14 days
(2
measurement
s: 1 work
and 1 nonwork day)
24h
Daytime
within-group
comparison
Enstrom et al.,
1996196
RCT
Night-time
24h
Reproducibility
of BP (between
the 2
measurements
over time)
Correlation with
clinical
diagnosis of
ICH
Reproducibility
of ICH
diagnosis
(repeated
ABPM
measurements)
Reproducibility
of ABP, BP
variability and
BP reduction
Reproducibility
on work and
non-work days:
SD;
reproducibility
over time (2
measurements,
There was no change in diurnal BP
variations.
This indicates that the short term
reproducibility of diurnal changes
in BP in the early phases of
untreated essential HT, is overall
satisfactory.
Classification of ICH based on a
single ABPM (using cut-offs
suggested in major HT guidelines)
has limited short-term
reproducibility
Repeated ABPM measurements at
a short time interval should be used
to ensure correct diagnosis of ICH
and improve CV risk stratification,
allowing a more appropriate
treatment strategy
Reproducibility of ABP levels and
BP varaiblity was fairly good.
Reproducibility of BP reductions
was fairly good for ABP levels, so
a single ABPM before and during
treatment is acceptable in a drug
intervention trial.
BP was higher during the work
day.
Daytime and night-time: there was
a SS difference in BP measurement
between the 2 readings
There was NS difference for night-
Hypertension (partial update)
Cuspidi et al.,
2002150
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
73
post-hoc analysis
(DIDIMA study)
Hermida et al.,
2002271
Case-series
1004
ABPM
sessio
ns
(529
studies
)
Borderline
HT,
suspected
WCH,
suspected
hypotension,
MHT, Tx
resistance, aHT treatment
OSC
538
HT
OSC
24h
3
readings
/hr
(daytim
e)
All: 20 min intervals
2 weeks apart)
time BP between the 2 readings
There were no major differences in
reporducibility if 1, 2 or 3
recordings / hour were used.
Arbritrary dividing lines for
day/night or according to patients’
own statement did not have any
major effect on the result.
But it may be wise to perform
recordings not less than every 30
mins for patients
D – 20 mins (6am – 6,
8 or 10pm)
Correlation of
shorter ABPM
periods with
24h ABPM
After excluding the first hour,
correlations for mean SBP the
subsequent 3-, 5- and 7-hour
periods demonstrated greatest
improvement in correlation when
session is increased from 4 to 6
hours.
6-hour ABPM can approximate the
overall mean BP obtained from full
24-hour ABPM.
Shortened sessions do not
characterise the influence of
circadian variation over the 24hour mean BP and may
overestimate 24-hour BP levels.
Comparison of
variations in BP
BP is significantly increased by the
novelty of wearing an ABPM
device for the first time (the
‘ABPM effect’).
Pressor effect remains statistically
significant for the first 6-8h of
monitoring, independent of gender,
N – 30 mins (6, 8 or
10pm – 6am)
2
readings
/hr
(nighttime)
48 h
48h
D – 20 mins (07.0023.00)
N – 30 mins (23.0007.00)
ABPM started on a
Hypertension (partial update)
Ernst et al., 2008200
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
weekday (Mon, Wed or
Fri)
74
Hernandez-del Rey
et al., 2007272
Historical caseseries
611
HT
OSC
48h
24h /
48h
Night and day defined
based on patient’s diary;
at least 14
measurements during
period of activity and at
least 7 during period of
rest
Recording intervals
(minutes between
measurements) not
given
day of the week of monitoring and
number of a-HT drugs used.
Differences between successive
days of ABPM are no longer
significant when patients were
evaluated for second or successive
times.
ABPM for just 24 h may be
insufficient for a proper diagnosis
of HT, evaluation of treatment
efficacy and identification of
dipping status in relation to targetorgan damage.
Reproducibility
of BP dipping
pattern in 24-h
vs 48-h ABPM
The percentages of patients
classified as non-dipper for the first
24 h, the second 24 h and the 48 h
average were 47, 50 and 48%
respectively.
When the first and second 24-h
periods were compared, 147 (24%)
subjects switched from dipper (D)
to non-dipper (ND) or vice-versa.
When the first 24-h period was
compared to the 48-h average, 66
(11%) subjects switched patterns.
The proportions were similar
separately for SBP and DBP, and
between treated and untreated
patients.
In subjects with poor ABPM
reproducibility, night-to-day ratios
were of an intermediate value
between those of subjects always
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
classified as Dipper or non-dipper.
Categorisation of D or non-dipper
based on a single 24-h ABPM is
moderately reproducible, since one
out of every five patients change
profile over the following 24 h.
A more reliable classification of the
BP circadian profile should be
performed by repeating a second
ABPM within a short period, but
the use of 48-h ABPM in clinical
practice should be assessed
according to cost-effectiveness
criteria.
75
Lede et al., 1997353
49
case-series
Mancia et al.,
1992386
case-series
29
Pregnant
women with
preeclampsia
(DBP≥90mm
Hg and
proteinuria
>300mg).
AUS
HT
AUS
24h
24h
3 different frequencies
of monitoring (FoM)
readings/ hour:
Similarities in
BP
measurements
between 3 FoMs
High FoM = 7/hr
Low FoM = 1/hr
Medium FoM = 2/hr
4 weeks
(2
measurement
s4 weeks
apart)
24h
Daytime (15 mins)
Night-time (20 mins)
24h
Reproducibility
of BP (between
the 2
measurements
over time; and
hourly vs mean
24h, SDD)
BP was similar in the three FoMs
studied at daytime and night-time.
There is therefore no strong
argument to perform ABPM at high
FoM
BP measurement at a lower FoM
may be better for the patient and
reduce equipment deterioration
whilst providing equivalent
information as supplied by a high
FoM
The second ABPM recording was
lower but was NS different from
the first
Reproducibility was lower for
hourly rather than 24h average BP.
This suggests that ABPM
measurement loses its advantages
for reproducibility if results are
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
analysed over hourly periods
Mancia et al.,
2004387
6000
HT (treated)
AUS or
OSC
1 week – 36
months
-
Night-time: not given
24h: not given
SR / MA of 44
trials
76
Mansoor et al.,
1994 389
Daytime: not given
25
HT
AUS
and
OSC
Mean 23
months
24h
Daytime
Night-time
within-group
comparison
24h
All: 15 min intervals
Mar et al., 1998390
138
HT (newly
diagnosed)
OSC
Not given
24h
within-group
comparison
Daytime (20 mins)
Night-time (1 hr)
24h
Murakami et al.,
2004416
within-group
135
General
population
(HT and NT)
OSC
7 days
-
Fitted on Thursday
between 10am – 2pm; D
- every 30 mins (0700 to
2200 hours)
Change in BP
response by
different
measurement
methods
Reproducibility
of BP (between
2 repeated
studies and over
time): SDD, coefficient of
variance and %
of people within
10mm and 5mm
SBP and DBP
Treatment-induced reduction in BP
is smaller for the night-time than
daytime average BP
The effect of anti-HT treatment is
unevenly distributed between day
and night
Results advocate a more systematic
adoption of ABP monitoring in
trials assessing CV protection by
anti-HT drugs
24h and night-time BP had better
reproducibility than daytime BP
(between studies and between
readings over time)
Diagnostic
accuracy with
varying number
of
measurements
Increasing the number of
measurements led to a reduction in
diagnostic error due to random
variability of BP.
Comparison of
weekly
variations in BP
Monday surge in BP was found in
the awake and morning BP but not
in the asleep BP
Morning BP surge on Monday was
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
comparison
N - 60 mins (2200 to
0700 hours).
Musso et al.,
1997420
40
NT
OSC
case-series
77
Octavio et al.,
2010456
450
Suspected
arterial HT
not
specifie
d
3 months
(4
measurement
s
each 28 days
apart)
24h
24h
24h
Daytime (15 mins)
Night-time (30 mins)
24h
Group
within-group
comparison
BP
readin
g
interv
al
Reproducibility
of BP (between
the 4
measurements
over time)
Reliability of
conventional vs
time-weighted
quantification of
24-h ABP
Day
(0600
2300)
Night
(2300
0600)
I
15
min
30
min
II
15
min
20
min
III
30
min
30
min
higher than on the other days of the
week except for Tuesday
Morning BP surge on a Monday
may be in accord with clinical
evidence that CV events more
frequently occur in the morning on
Monday
There was high agreement between
the 4 readings
BP values were lower during the
4th reading (vs 1st)
People should not be labelled as
HT based on initial readings, since
initial ABPM may yield higher
values than later monitoring
Higher number of readings per
hour during daytime leads to an
overestimation of conventional 24h average BP, particularly in
individuals with preserved
nocturnal BP dipping.
This can be avoided either by
scheduling the same number of
readings/h throughout 24 h or by
performing a time-weighted
quantification of 24-h BP
The clinical implications of these
different approaches deserve
further investigation.
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
6461
ISH or high
DBP
OSC
3 months
2 (3
months
apart)
Daytime (10 mins)
Night-time (30 mins)
case-series
Reproducibility
over time (2
measurements,
3 months apart)
24h
78
Schillaci et al.,
1994527
24
HT
OSC
case-series
1 week
(2
measurement
s1 week
apart)
24h
Daytime (15 mins)
Night-time (15 mins
session 1, 1hr session 2)
Reproducibility
of BP (between
the 2
measurements
over time)
24h
Schwartz et al.,
2000530
within-group
comparison
143
NT
AUS
1 week
24h
Active period (daytime)
Inactive period (nighttime)
All: 10 min intervals
Intraindividual
BP variability
(SDs), during
the active
(daytime) and
inactive
(nighttime)
periods of the
Small but SS decreases in average
daytime BP / no change in average
nighttime BP occur when ABPM is
performed twice 3 months apart.
There was a SS increase in SBP
when the period between midnight
and 5 am was considered in
nighttime analysis.
ABPM shows better reproducibility
than office BP, particularly for 24h
BP. Nighttime BP was less
reproducible than daytime BP,
probably due to sleep disturbance
which was reported in 2/3 of
patients.
There was NS difference in
daytime or night-time systolic or
diastolic BP and heartrate between
the two sessions
A low number of cuff
measurements of BP during the
night (1 per hour) provides similar
results to a high number of
measurements in terms of sleep BP,
and changes of BP from wake to
sleep.
Men: had greater BP variation
(SBP and DBP) during the inactive
period (vs. active period)
Women: SBP – there was NS
difference in BP variation during
the inactive period (vs. active
period). DBP – as for men.
Hypertension (partial update)
Palatini et al.,
1994473
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
day
Schwartz et al.,
2000531
240
NT
AUS
1 week
24h
Active period (daytime)
Inactive period (nighttime)
within-group
comparison
All: 10 min intervals
Sheps et al.,
1994538
294
HT and NT
AUS
2 months
(2
measurement
s2 months
apart)
24h
Daytime (7.5 mins) and
other time frequencies
56
unclear
OSC
7 days
7 days
of 24h
recordin
gs
Daytime (30 mins)
within-group
comparison
Intraindividual
BP variability
(SDs), during
active (daytime)
and inactive
(nighttime)
periods of the
day
Reproducibility
of BP (between
the 2
measurements
over time):
79
Shinagawa et al.,
2002541
case-series
Night-time (1 hour)
BP variability
on different
days of the
week
24h
Stenehjem et al.,
2004562
within-group
comparison
75
HT
AUS
4 weeks
measurement
s before and
after 4 week
observation
period (2
separate
work days)
24h
D – 20 mins (0700 –
2200)
N – 30 mins (2200 –
0700)
Reproducibility
of BP
variability,
white coat effect
and dipping
pattern
Men and women: there was greater
BP variation (SBP) during the
inactive period (vs. active period)
Women: DBP – there was NS
difference in BP variation during
the inactive period (vs. active
period)
As few as six hours of monitoring
with two to three readings/hour
achieved most of the gain in
precision obtainable by going from
single BP readings toward
continuous measurement during an
entire awake period
The average SBP (daytime) is
higher on the first day of
monitoring vs the other 6 days.
Daytime BP was lowest on
Sundays and the day-night ratio
was optimal on weekends.
Average ABPs are highly
reproducible in patients with
uncomplicated essential HT of
limited duration.
Nocturnal dipping pattern also
reproduced satisfactorily.
White coat effect and variability
are greatly attenuated during
repeated measurements, and these
measures may thus be of less utility
in clinical practice.
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
within-group
comparison
Suarez et al.,
2003573
80
133
HT
(untreated)
OSC
2 work days
24h
Every 20 mins
Test-retest
variability
(correlations
and SDD)
261
HT
OSC
24h
24h
D – 20 mins (07002400)
Agreement
between ABP
daytime average
and 24-h
average for
diagnosing HT
and assessing
effects of antiHT treatments
(sensitivity /
specificity)
In 90% of the records there was
agreement between both criteria
Daytime and 24 h average BP may
carry similar information for
diagnosing HT and assessing the
effects of anti-HT treatment in
clinical practice.
ABPM used only during the
daytime could be better tolerated
and agreed to by patients than 24 h
monitoring.
Consistency
(median
differnce
between the 2
recordings);
repeatability (2
x SD of the
changes
between the 2
recordings)
Reproducibility
24h and Daytime ABPM was better
than night-time BP (all were better
than clinic)
N – 30 mins (2400 –
0700)
retrospective
diagnostic caseseries
Reference standard:
mean 24h ABP
(≤125/80)
Thijs et al., 1992595
102
ISH
OSC
within-group
comparison:
substudy of SystEur trial
1 month
(2
measurement
s – 1 month
apart)
24h
Index test: mean awake
ABP (<135/85)
Daytime
Night-time
24h
All intervals not <30
mins
Trazzi et al.,
ABP and pulse pressure and of
nocturnal fall in BP have the most
prognostic relevance and are of
great value in clinical practice.
Mean 24h (was better than awake
or asleep BP)
34
HT
AUS
4 weeks
24h
Daytime (10 mins)
There WAS NS differnce in SBP /
Hypertension (partial update)
Stergiou et al.,
2002563
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
1991600
(2
measuremnts
– 4 weeks
apart)
case-series
Van der Steen et
al., 1999608
45
HT
within-group
comparison
Van Ittersum et al.,
1995609
20
HT and
WCH
Night-time (20 mins)
24h
AUS
device
may not
be truly
ABPM
2-3 weeks
(2
measuremnts
– 2-3 weeks
apart)
24h
OSC
24h
24h
Night-time (30 mins)
24h
81
Retrospective
comparative study
with historical
control
Daytime (15 mins)
Night-time (20 mins)
long fixed sleep period:
waking 7am-10pm and
sleeping 10pm-7am
short fixed sleep period:
waking 10am to 11pm
and sleeping 1am-7am
pts diary sleep period:
actual sleep times
retrospective caseseries
Wallace et al.,
2005622
Daytime (15 mins)
31
HT
AUS
2 separate
weekdays, 23 days apart
SAME
group:
monitoring
began at
24h
24h
SAME group: first
reading 177-1900; OPP
group: sessions
randomised to begin in
morning (0700-0900) or
evening (1700-1900).
D - 15 ± 5 minutes
of BP (between
the 2
measurements
over time)
DBP measurements 4 weeks apart
(24h ABPM)
24h ABPM was more reporducible
than office BP due to a larger
number of measurements.
Reproducibility
of BP (between
the 2
measurements
over time)
There was poor reproducibility.
24h and daytime BP were better
than night-time measurements.
Differnce in BP
using long and
short sleep
periods vs
actual sleep
period (pts
diary)
A short sleeping period gives
accurate measures of blood
pressure during sleep.
The long sleeping period method
should be avoided as it can
overestimate BP during sleep.
Reproducibility
of BP variables:
averages, 24-h,
day-time, nighttime, crest,
trough,
trough:crest
(Intra-class
For SBP the ABPM was only
reproducible when monitoring
began at the same time of day and
not when variables were measured
at opposite times of day
TrBP and average 24-h SBP were
significantly higher when the
monitoring session began in the
Hypertension (partial update)
Frequency of measurements
Diagnosis of Hypertension
Reference / study
type
Frequency of measurements
same time of
day
Zakopoulos et al.,
2001654
25
HT
OSC
OPP group:
sessions
randomised
to begin in
morning or
evening
4 months
(0600-2200)
N - 30-45 ± 5 minutes
(2200-0600)
24h
Four times
(four(interval
s of 1 week
each)
case-series
correlation)
Daytime
Reproducibility
over time (2
measurements,
2 weeks apart)
Night-time
morning compared with the
evening
Reproducibility of DBP was
similar between SAME and OPP
conditions.
Ambulatory BP variables were
consistently higher when
monitoring session began in the
morning
There was no difference between
the 4 readings (over time) for 1h,
24h daytime or night-time (SBP or
DBP)
24h
82
All: 15 min intervals
and 1 hr intervals
NT = normotensives; HT = hypertensives; ICH = isolated clinic HT; AUS = auscultatory device; OSC = oscillometric device; D = daytime; N = night-time; TrBP = trough BP.
Table 23: Day and night intervals and results for prognostic studies assessing the optimal ABPM protocol
N
Follow-up time
Day protocol (mins)
Night protocol (mins)
Best: day (D), night (N) or 24h
Hansen et al., 2005253
1700
Up to 9.5 years
15
30
D + N + 24h
326
5682
Median 9.5 years
15, 20, 30
20, 30, 45
All intervals are the same. D + N
+ 24h
688
Mean 9.2 years
60
60
D + N + 24h
324
Mean 51.5 months
30
30
N + 24h
DAY and NIGHT and 24h
Kikuya et al., 2007
Khattar et al., 2001325
NIGHT and 24h
Suzuki et al., 2000576
DAY and 24h
Update 2011
Reference / study type
Hypertension (partial update)
Diagnosis of Hypertension
Reference / study
type
Follow-up time
Day protocol (mins)
Night protocol (mins)
Best: day (D), night (N) or 24h
256
Mean 84 months
15
15 or 30
Morning was as good as D + 24h
2232
Median 5 years
30
<60
D + 24h
7458
analysed
Median 9.6 years
15-30
30-60
D+N
741
Mean 7.4 years
15
30
D+N
8945
Mean 5.8 years
15-30
15-30
D+N
872
Mean 6.6 years
20
20
D+N
7458
Median 9.6 years
-
-
D+N
1542
Mean 10.6 years
30
30
D+N
Fagard et al., 2005211
391
Median 10.9 years
15
30
N
210
302
Median 6.8 years
15-30
30-60
N
2051
Mean 10.9 years
20
20
N
951
Up to 9.1years (mean
range 0.1 – 11.4
years)
20 or 30
30 or 60
N
837
Mean 4.4 years
≤30
≤30
N
5292
Mean 7.9 years
30
30
N
Gosse et al., 2001
Clement et al., 2003
131
DAY and NIGHT
Boggia et al., 200788
Cipriano and Gosse et al., 2001237
Pickering et al., 2007
491
Bjorklund et al., 2004
Li et al., 2008
77
363
Metoki et al., 2006405
NIGHT
Fagard et al., 2008
83
Sega et al., 2005
534
Ingelsson et al., 2006
284
Staessen et al., 1999557
Dolan et al., 2005
178
D = daytime; N = night-time
Hypertension (partial update)
N
237
Diagnosis of Hypertension
Reference / study type
N
Follow-up time
Day protocol (mins)
Night protocol (mins)
Best: day, night or 24h
25
4 months
15
15
D + N + 24h
45
2-3 weeks
15
30
D + 24h
261
24h
20
30
D + 24h
102
1 month
≥30
≥30
D + 24h
DAY and NIGHT and 24h
Zakopoulos et al., 2001654
DAY + 24h
Van der Steen et al., 1999608
Suarez et al., 2003
Thijs et al., 1992
573
595
NIGHT + 24h
Palatini et al., 1994473
6461
3 months
10
30
N + 24h
Mansoor et al., 1994
389
25
Mean 23 months
15
15
N + 24h
Antivalle et al., 1990
46
22
4 weeks
-
-
N + 24h
Schillaci et al., 1994527
24
1 week
15
15 or 60
D + N (60minswas fine for night)
Schwartz et al., 2000530
143
1 week
10
10
D
531
240
1 week
10
10
D
52
64
1-2 years
30
60
D
72
2 years
15
30
24h
133
2 work days
20
20
24h
1004 sessions
24h
20
30
6h ≈ 24h
Hermida et al., 2002271
538
48 h
20
30
>24h
Calvo et al., 2003111
823
48 h
20
30
>24h
DAY + NIGHT
84
DAY
Schwartz et al., 2000
Asagami et al., 1996
≤24h
Campbell et al., 2010114
Stergiou et al., 2002
Ernst et al., 2008
563
200
>24h
OTHER – INTERVALS SPECIFIED
Sheps et al., 1994538
294
2 months
7.5, 20 or 30
-
20 and 30 mins are almost as good
(for D)
Lede et al., 1997353
49
24h
7.5, 30 or 60
7.5, 30 or 60
All times are similar
Hypertension (partial update)
Reference / study type
Diagnosis of Hypertension
Table 24: Day and night intervals and results for reliability/reproducibility studies assessing the optimal ABPM protocol
Follow-up time
Day protocol (mins)
Night protocol (mins)
Best: day, night or 24h
29
4 weeks
15
20
24h was better than hourly
456
450
24h
15 or 30
20 or 30
D had lower readings,or perform
the same number of readings for
24h
80
14 days
20
20
20, 30 or 60 mins are fine
138
Not given
20
60
Increased measurements are better
100
1 month
30
-
More day measurements are better
34
4 weeks
10
20
-
Octavio et al., 2010
Enstrom et al., 1996196
Mar et al., 1998
390
Coats et al., 1992
133
NOT SPECIFIED
Trazzi et al., 1991600
Van Ittersum et al., 1995
609
20
24h
15
20
-
Cuspidi et al., 2002150
208
3 weeks
15
20
-
151
611
1-4 weeks
15
20
-
30
1 month
15
30
-
31
2-3 days
15
30-45
-
75
4 weeks
20
30
-
43
2 weeks
30
30
-
56
7 days
30
60
-
135
7 days
30
60
-
6000
1 week – 36 months
-
-
-
40
3 months
15
30
-
611
48h
-
-
-
Cuspidi et al., 2007
Asmar et al., 2001
56
Wallace et al., 2005
622
85
Stenehjem et al., 2004
562
Eguchi et al., 2010190
Shinagawa et al., 2002
Murakami et al., 2004
Mancia et al., 2004
Musso et al., 1997
541
416
387
420
Hernandez-del Rey et al.,
2007272
+ = ‘or’ ; D= daytime; N = night-time
Hypertension (partial update)
Mancia et al., 1992
N
386
Diagnosis of Hypertension
Reference / study type
Hypertension (partial update)
Diagnosis of Hypertension
6.4.1.2
Health economic evidence
No relevant economic studies were identified relating to ABPM measurement protocols.
6.4.1.3
Evidence statements – clinical
The 17 prognostic studies recommend the following regimens (as the best predictors of CV
events) :
 All day measurements are needed (11 studies):
o day and night– day and night measurements predict different outcomes (four
studies)88,363,405,491
o 24h, day and night were all good predictors of outcome (five studies)77,237,253,325,326
o 24h and day were the best predictors of outcome (one study)131
o 24h and night were the best predictors of outcome (one study)576
 Night BP only is sufficient (a good predictor of outcome) (six studies)178,210,211,284,557534
 A single BP measurement on rising is sufficient – this is as good as using the 24h or
daytime mean for predicting outcome (one study)237
 Excluding the first two hours does not improve accuracy (one study)557
 SBP is sufficeint (a good predictor of outcome) but DBP is not (four studies: one study SBP in >60 years, DBP<60 years)77,237,325,534
 DBP is sufficient (a good predictor of outcome) but SBP is not (two studies: one study SBP in >60 years, DBP<60 years)253,325
86
Update 2011
The 36 reliability/reproducibility studies showed the following:
1. The optimum interval between measurements:
 Repeat ABPM over a short time interval (one study)151
 A greater number of readings/hr leads to an overestimation of BP: use the same number
readings over 24 hours or use a time-weighted calculation of 24h BP (one study)456)
 One reading per hour for night-time is equivalent to a 15 min interval for night-time BP
(one study)527
 A short sleep period (1-7am) is more accurate than using a long sleep (10pm – 7am) (one
study)609
 Daytime BP: taking more measurements is better than just one measurement (one study)133
 More measurements taken lead to less diagnostic error (one study)390
 Taking 2-3 readings/hr for 6 hours is almost as good as continuous measuring every 7.5
mins for daytime ABPM (one study)538
 There is no difference between taking 1, 2 or 3 recordings per hour, but using an interval of
<30 mins is probably not so good for the patient (one study)196
 There was no differnce between taking one, two or seven recordings per hr. However a
lower number of recordings is probably better for the patient and for the longevity of the
equipment (one study)353
2. When to begin measurements:
 SBP – take measurements at the same time of day, not at opposite times (one study)622
 Mean 24h BP is higher if measurements are started in the morning rather than the evening
(one study)622
 DBP – readings are not affected by the time of day that measurements are taken (one
study)622
Hypertension (partial update)
Diagnosis of Hypertension
5. What day of week to perform ABPM:
 Monday morning BP surge is greater than on other days (one study)416
 The day of the week does not affect the pressor effect ie. lower BP values are obtained
with consecutive measurements (two studies)111,271
 Daytime BP is lowest on Sunday; the optimal day-night ratio occurs on weekends (one
study)541
• BP is higher on a work day (one study)196
6.4.1.4
Evidence statements – economic
 No relevant cost-effectiveness evidence was identified.
87
Update 2011
3. The best time of day to take measurements
 All day measurements are needed (16 studies):
o One hour (one study), 24h, day, night (two studies)150,654
o Day and night are best (two studies)387,527
o Day and 24h are best – one study showed 24 hour BP was slightly better but using 6
hour BP was sufficient if patients are not able to tolerate / comply with 24 hours of
measuring (four studies)473,573,595,608
o Night and 24 hour measurements gave greater reproducibility (two studies)46,389
o Daytime measurements are best (especially for men in one study; three studies)52,530,531
o Mean 24 hour measurements are best (two studies)114,563
o 24h BP is similar to 6 hour BP: but 6 hour BP may overestimate the value as it does not
account for 24 hour BP variation (one study)200
4. How often to repeat measurements (over time)
 Twice - four weeks apart: there was decreased variability and WCH (one study)562; similar
measurements were found at both times (one study)600
 Twice - two weeks apart (one study)190
 Twice (second) or successive times, or 48 hours – this accounts for: circadian variation, the
ABPM effect (higher BP the first time ABPM is used), the pressor effect (lower BP
readings achieved with consecutive measurements) - three studies111,271,272
 Four times (four weeks apart): there was high agreement between the measuerments but
the fourth measurement gave a lower BP reading – therefore don’t label someone as being
HT on the basis of an initial ABPM (1 study)420
 Twice (three months apart): BP was SS lower in the day but not at night or over 24h BP
measurement (one study)473
 The first day of monitoring gave higher BP readings than measurements of the other six
days (one study) 541
Hypertension (partial update)
Diagnosis of Hypertension
6.4.2
6.4.2.1
Home blood pressure measurement
Review question: In adults with primary hypertension, what protocol should be used when
measuring blood pressureat home for treatment and diagnosis?
Clinical evidence
6.4.2.2
Economic evidence
No relevant economic studies were identified relating to HBPM measurement protocols.
6.4.2.3
Evidence statements – clinical
The studies showed the following:
The optimum number of readings to take (seated)
 Only one reading is sufficient (two studies)123,283
 Two or >two readings are needed: (two studies) 203,302
 Three readings are needed: (two studies)191,612
88
Update 2011
The literature was searched for all years and studies published since the original guideline
(2003 onwards) were included. All study types were included, if the population did not
consist of people who were exclusively diabetic or had CKD. Validation studies of home
blood pressure machines were excluded.
Eight studies53,191,203,302,315,316,464,565,611,612 were found that fulfilled the inclusion criteria and
assessed what protocol should be used when measuring home BP in for the treatment and
diagnosis of adults with primary hypertension. Two of the studies (1 study;53,464 one
study315,316) were each published as two separate papers reporting different assessment
methods or outcomes, so these studies have only been counted once, however results from
both papers are reported and referenced here.
The studies addressing the question were categorised into two different types:
 Prognostic studies (two studies; three papers)53,53,565 – those that assess the prognostic
significance of home blood pressure and the optimal schedule for measurement based on
outcome data
 Reliability / reproducibility studies (seven studies; eight papers)191,203,302,315,316,565,611,612 those that assess any of the following - the optimal home blood pressure schedule based
on:
o the reproducibility of home blood pressure
o its stability over time
o its relationship (correlation) with ABPM values
o its ability to identify people diagnosed with Hypertension / Normotension
o its ability to identify treatment responders
Reliability /repeatability studies were deemed to be applicable to the question because they
showed which aspects of the HBPM protocol were the most reliable, and therefore served as
an indication of the ‘best’ / optimal HBP measurements to be taken.
All prognostic studies were found to be methodologically sound / have a low risk of bias (see
quality assessment summary tables in appendix F).
Details of all the studies are included in Table 25 and Table 26. NOTE: all home blood
pressure measurements in the studies were taken when the patient was seated.
NOTE: For the prognostic studies, the ‘best method’ was chosen as the method of measuring
BP that best predicted (ie. statistically significant predictors and higher HR values) clinical
outcomes (after adjustment for covariates in multivariate analyses). For the
‘reproducibility/reliability studies’ the ‘best method’ was chosen as the the method / protocol
of measuring blood pressure that was the most reliable or repeatable.
Hypertension (partial update)
Diagnosis of Hypertension
The optimum interval between measurements
 Take a one minute interval, not every ten seconds (one study)191
Should any readings be discarded?
 The first and second reading are both fine (one study)565
 Discard the first reading (three studies, four papers) 315,316,565,568
 Discard day one readings (one study)565
 Discard day one readings (two studies) 565,568
 Keep day one readings (one study)302
 Discard day one and daytwo readings (one study)612
The best time of day to take measurements
 Morning and evening are best (two studies, three papers)53,464,565
 Morning only is sufficient (one study)283
 Morning and evening are best (one study) 302
How many days to take measurements
 Three days (four studies)123,228,283,568
 Four or more days (one study)302
 Five or more days (two studies)203,612
 Seven days (one study, two papers) 315,316
89
Reference / study type
N
Population
Device
Consecutive
readings
Days
Time of measurement
Outcomes
Stergiou et al.,
2010565
665
HT
AOD
2
3
M – seated, after 5 mins
rest
E – seated, after 5 mins
rest
CV events
(fatal / nonfatal)
Within-group
comparison (DIDIMA
STUDY)
90
Ohkubo ey al., 2004
and Asayama et al.,
200653,464
Within-group
comparison
(OHASAMA
STUDY)
1766
General
population
(HT and NT)
SOD
≥2
4 weeks
M – seated, within 1hr
waking
E – seated, just before
going to bed
Stroke
Proposed protocol (authors’
conclusions) – best
prognostic ability
more readings averaged
(from 1-12) increased the
prognostic ability.
Take the 1st or 2nd readings;
morning or evening are
equally good; discard 1st day
Morning and evening are
equally good; there is no
threshold (1-14
measurements) – but take as
many measurements as
possible (preferably >14
measurements)
NT = normotensives; HT = hypertensives; AOD = automatic oscillometric device; SOD = semiautomatic oscillometric device; E = evening; M = morning; MS = mercury
sphygmomanometer
Hypertension (partial update) Update 2011
Frequency of measurements
Diagnosis of Hypertension
Table 25: Study details and overall results for prognostic studies assessing the optimal home blood pressure protocol
Table 26: Study details and results for reliability/reproducibility studies assessing the optimal home blood pressure protocol
Frequency of measurements
Reference / study
type
N
Population
Device
Consecutive
readings
Days
Time of measurement
Mathematical
method
Proposed number of
measurements (authors’
conclusions)
Verberk et al., 2005611 MODERATE QUALITY systematic review of 4 within-group comparison observational studies (studies below)
SR study 1:
Celis et al., 1997123
74
Elderly HT
MS
1
M – lying in bed
M – after 10 mins
standing
E – standing before
going to bed
E – lying in bed for 10
mins
Variability
(SD); t-test
3
workdays
M (6 – 10am)
E (5 – 11am)
Test-retest
variability (SD),
correlation with
ABPM
Take the average of the 2nd
and 3rd working day
100
91
SR study 2:
Stergiou et al.,
1998568
Within-group
comparison
SR study 3:
Garcia-Vera et al.,
1999228
Within-group
comparison
SR study 4:
Imai et al., 1993283
Within-group
189
HT
AOD
2
Update 2011
Within-group
comparison
Take one reading / day for 3
consecutive days
48
HT
SOD
1
8
M
E
At work
Test-retest
variability (SD),
Generalisability
theory
Take one reading at work
and one at home for 3
consecutive days for reliable
estimates for 2 months
871
NT and HT
SOD
1
28
M - <1h after awakening
Variability (SD)
Take one reading/day in the
morning for 3 consecutive
days
Hypertension (partial update)
Diagnosis of Hypertension
Reliability / reproducibility studies
Other studies
Stergiou et al.,
2010565
Within-group
comparison
(DIDIMA STUDY)
Kawabe et al., 2005
and 2008315,316
665
HT
AOD
2
3
M – seated, after 5 mins
rest
E – seated, after 5 mins
rest
Variability (SD)
More readings averaged
reduced variability (from 112); discard the first day (as
this gave unstable values)
700
General
population
(HT and NT)
SOD
3
7
M – seated, within 1hr
waking (before breakfast
and medication, after
urination)
E – seated, before bed
(not within 30 mins
bathing)
Correlation with
clinical
diagnosis of HT
/ NT
Take 7 day measurements
for diagnosis (more
pronounced using 1st vs.
mean 2nd and 3rd
measurements or evening
BP): this led to a diagnosis
of HT more frequently, and
NT less frequently
57
Known or
suspected HT
AOD
3
8 weeks
(4days/
week)
M – 10sec or 1 min
intervals (randomised to
eaither)
E - 10sec or 1 min
intervals (randomised to
either)
Correlation with
ABPM and
Office BP
Take a 1 min interval of 3
measurements (this gave a
better estimate of average
daytime ABPM level; 10sec
intervals gave higher
readings than 1 min)
464
HT
AOD
2
7
M – 1-2 min intervals
E – 1-2 min intervals
Correlation with
ABPM
Within-group
comparison
92
Eguchi et al., 2009191
Cohort study
Johansson et al.,
2010302
Cohort study
Ewald et al., 2006203
Mean number 27.5
53
HT
AOD
≥1
12 weeks
M
Identification of
Take duplicate
measurements, at least 4
days (evening and morning);
don’t discard 1st day
measurements (there was
NS difference in correlation
with ABPM when the 1st
day was excluded)
Take at least 2
Hypertension (partial update)
comparison
Diagnosis of Hypertension
Frequency of measurements
treatment
responders
(sensitivity/
specificity);
response to
Treatment
measurements/day (this
gives a better response to
treatment); take at least 5
readings/week (this was the
threshold for correctly
predicting response to
treatment)
M – seated, after 5 mins
rest (1 min interval
between measurements)
E – seated, after 5 mins
rest (1 min interval
between measurements)
Correlation with
ABPM
Take a minimum of 5 days;
3 consecutive morning and
evening measurements;
discard 1st two days and 1st
reading of each triplicate
(for calculating mean
values) – this is a time
consuming protocol, so use
it for a decision to start or
change treatment, or for
special patient groups
Post-hoc analysis of
RCT (OLMETEL
STUDY): thus cohort
Verberk et al.,
2006612
Post-hoc analysis of
RCT (HOMERUS
STUDY) thus cohort
216
HT
AOD
3
7
93
NT = normotensives; HT = hypertensives; AOD = automatic oscillometric device; SOD = semiautomatic oscillometric device; E = evening; M = morning; MS = mercury sphygmomanometer
Hypertension (partial update)
E
Diagnosis of Hypertension
Frequency of measurements
Hypertension (partial update)
Diagnosis of Hypertension
6.4.2.4
Evidence statements – health economic
 No relevant cost-effectiveness evidence was identified.
6.5 Link from evidence to recommendations
94
Update 2011
Clinic blood pressure measurement (CBPM) on repeated clinic visits has long been the
standard method for the diagnosis of hypertension and subsequent monitoring blood pressure
control on treatment in clinical practice. The increased availability of automated blood
pressure measuring devices has led to their increased use in clinical practice and clinical
studies. Home blood pressure measurement (HBPM) or ambulatory blood pressure
measurement (ABPM) both provide multiple measurements of blood pressure away from the
clinic setting in a more usual environment.
This raised the question as to whether ABPM and/or HBPM may provide better prognostic
information with regard to the relationship between blood pressure and clinical outcomes. The
predictive value for clinical outcomes of blood pressure measurement based on clinic blood
pressure measurement (CBPM), home blood pressure measurement (HBPM) and ambulatory
blood pressure measurement (ABPM) were compared. Three pooled analyses were identified.
210,254,326
The clinical outcomes of interest were mortality, stroke, MI, heart failure, diabetes,
vascular procedures, hospitalisation for angina, and other major adverse cardiac and
cerebrovascular events (MACCE). All other studies identified were observational and
comprised nine prognostic studies77,159,178,210,253,254,284,326,404 that compared CBPM with
ABPM, five studies86,211,438,534,564 that compared CBPM with HBPM and two studies211,534
that compared all three methods for blood pressure measurement. The studies included adult
patients with normal blood pressure, suspected hypertension and known hypertension across a
wide age range (30 to 71 years). All of the studies were deemed to have a low risk of bias.
The results of this analysis showed that when CBPM was compared to ABPM in 8 out ofthe
9 studies77,159,178,210,253,254,284,404 ABPM was superior to CBPM at predicting clinical events
there was no difference in one study. 326 ABPM can also provide data on the 24 hour average
BP, daytime average BP and night-time average BP. The GDG noted that in some studies the
daytime ABPM average was the most predictive of clinical outcomes, whereas in others the
ABPM night-time average was the most predictive but there was no conclusive evidence
suggesting a preference for day versus night-time averages. The GDG noted that from a
practical perspective, when comparing different methods, ABPM daytime averages are
preferred because they allow easier comparison with CBPM and HBPM averages which are
also usually taken during the daytime.
There was less data comparing CBPM with HBPM in only three studies. 86,438,564 HBPM was
superior to CBPM at predicting clinical outcomes in two of these studies86,438 and no
difference between the methods was noted in one small study. 564
All three blood pressure measurement methods were compared with each other in only two
studies in one of which there was no difference in their predictive value and in the other,
ABPM and HBPM were similar to each other but superior to CBPM at predicting clinical
outcomes.
Taken together, the GDG concluded that the analysis of these studies showed that CBPM was
never superior to ABPM or HBPM at predicting clinical outcomes. Furthermore, ABPM was
never inferior to other methods and was most often the best predictor of clinical outcomes.
HBPM also appeared superior to CBPM at predicting clinical outcomes but there was less
data with HBPM when compared ABPM. The GDG concluded that multiple blood pressure
measurements away from the clinic setting are the best predictor of blood pressure-related
clinical outcomes and that to date, studies with ABPM provided the most robust evidence.
The GDG considered the reasons for this and noted that this in part, could relate to the fact
that ABPM and HBPM are providing more measurements and more representative data of a
person’s usual blood pressure away from the clinic setting. It could also relate to the fact that
Hypertension (partial update)
Diagnosis of Hypertension
some people diagnosed as hypertensive based on their CBPM in reality have much lower
blood pressures according to their ABPM or HBPM averages, i.e. white coat hypertension or
a white coat effect, and consequently are at much lower risk of clinical outcomes than their
CBPMs suggest.
That said, the GDG felt that more prospective data from epidemiological studies and clinical
intervention trials, comparing the prognostic value of CBPM versus HBPM versus ABPM
should be undertaken to better inform this prognostic relationship and better define treatment
thresholds and targets according to daytime versus night-time averages and the optimal
protocols for HBPM and ABPM measurement.
As well as looking at prognostic studies the GDG reviewed studies that compared the
sensitivity and specificity of CBPM, HBPM and ABPM in order to address the important
question of which is the best method to measure blood pressure to diagnose hypertension. A
recent systematic review and meta-analysis 275 examined the relative effectiveness of CBPM
or HBPM versus ABPM for establishing the diagnosis of hypertension. ABPM was used as
the reference standard for this analysis on the basis that: i) it is a superior predictor of clinical
outcomes (see above); and ii) ABPM is the test resorted to in clinical practice when there is
uncertainty about the diagnosis of hypertension, thus, ABPM is the de facto reference
standard for confirming the diagnosis of hypertension in clinical practice. Thus, the GDG
agreed that it was appropriate to adopt ABPM as the reference standard for the analysis of the
three different BP monitoring modalities to establish the diagnosis of hypertension. This
systematic review included 20 studies (N=5863). For the purposes of the analysis, an ABPM
daytime average of 135/85mmHg was taken as the threshold for the diagnosis of hypertension
and the performance of CBPM or HBPM versus this reference standard was compared. The
CBPM and HBPM thresholds for diagnosis of hypertension were 140/90mmHg and
135/85mmHg respectively. Nine studies that used these thresholds were meta-analysed.
The meta-analysis found that, compared with ABPM, CBPM had a mean sensitivity of 74.6%
(95% CI, 60.7 to 84.8) and specificity of 74.6% (47.9 to 90.4) for the diagnosis of
hypertension and HBPM had a mean sensitivity of 85.7% (78.0 to 91.0) and specificity of
62.4% (48.0 to 75.0). Neither differences in sensitivity or specificity between HBPM and
CBPM were significant. In this context, “sensitivity” is the number of people who are
diagnosed with hypertension according to CBPM or HBPM as a proportion of all those who
actually have hypertension as defined by the ABPM reference standard. “Specificity” is the
number who test negative for hypertension according to CBPM or HBPM as a proportion of
all those that actually do not have hypertension as defined by ABPM. Thus based on the
specificity results from the primary analysis of the meta-analysis CBPM will misdiagnose
25% of people who do not have hypertension as hypertensive; with HBPM this figure is 38%.
In addition, based on sensitivity, with CBPM 25% of people with hypertension will
mistakenly be diagnosed as not hypertensive; with HBPM that figure is 14%.
However, the studies included in the meta-analysis for CBPM were in a range of populations
and a sensitivity analysis was also reported which included only studies with a mean BP close
to or above the diagnostic threshold. This is relevant because sensitivity and specificity vary
with disease prevalence – while it is often asserted that sensitivity and specificity are
independent of disease prevalence it has been demonstrated that when categorisation is based
on a continuous trait, as with hypertension, this is not the case. 98 In this analysis CBPM
sensitivity increased to 85.6% (CI 81.0 to 89.2) and specificity decreased to 45.9 (CI 33.0 to
59.3). The HBPM studies were all in this restricted population and so the analysis for HBPM
remained the same. With this restricted analysis CBPM and HBPM are virtually identical in
terms of sensitivity, but HBPM was now more specific than CBPM. This sensitivity analysis
was considered by the GDG to be more relevant to the guideline as screening the general
population is outside of its scope.
95
Hypertension (partial update)
Diagnosis of Hypertension
The GDG also considered a sensitivity analysis looking at the impact of the diagnostic
threshold on the performance of the different diagnostic methods. Perhaps not surprisingly,
the specificity of CBPM for diagnosing hypertension improved when the CBPM blood
pressure threshold for diagnosis is increased, i.e. those defined as hypertensive when their
CBPM is higher are more likely to be hypertensive according to ABPM. However, the
corollary was also true, i.e. that the accuracy of diagnosis of hypertension when comparing
CBPM with the ABPM reference standard is most uncertain in those who blood pressure is
close to the CBPM diagnostic threshold of 140/90mmHg.
This detailed analysis suggested that the current practice of using CBPM to define
hypertension will lead to drug treatment being offered to a substantial number of people who
are normotensive according to ABPM. The GDG recognised that these data have profound
implications for the diagnosis of hypertension. Firstly, they suggest that some patients
randomised and treated in clinical outcome trials on the basis of their CBPM, may not have
been hypertensive, potentially diluting and underestimating the true benefits of treatment in
those who were hypertensive. Secondly and perhaps more importantly, these findings suggest
that the current practice of using a series of CBPM alone for the diagnosis of hypertension can
lead to inaccurate diagnosis.
Screening for hypertension was outside the scope of this guideline. However, the GDG agreed
it is not practical to use ABPM or HBPM as a screening tool, despite them potentially
offering greater accuracy than CBPM. The working assumption was that CBPM would still be
used for screening patients and that the key decision that remained was how the diagnosis
should be confirmed.
Taking into account the prognostic data and the meta-analysis of sensitivity and specificity,
the GDG agreed that ABPM appeared to provide the best method of confirming a diagnosis of
hypertension. The GDG also considered that a change in practice as profound as this required
clear evidence that ABPM would not only be a more effective means of diagnosis but also, a
more cost-effective means of establishing the diagnosis of hypertension.
The GDG agreed the most practical method to diagnose hypertension would be to use CBPM
as a screening tool and that those people with a CBPM ≥140/90mmHg measured using the
recommended standardised conditions, should then be offered ABPM to confirm or refute the
diagnosis of hypertension based on a diagnostic threshold of an ABPM daytime average of
≥135/85mmHg.
The GDG reviewed the data regarding the number of measurements required to establish the
ABPM daytime average blood pressure. The number of measurements taken during
prognostic studies varied from every 15 minutes to every hour during the daytime. The GDG
concluded that two measurements per hour should be taken during normal waking hours, e.g.
08.00hrs to 22.00hrs and that a minimum of 14 readings should be used to derive the daytime
average blood pressure. This means that patients would not necessarily need to wear the
ABPM monitor for a full 24hrs, depending on the time the monitoring session was initiated.
For practical reasons and efficiency in use of the monitors, not every monitoring session will
begin at 08.00hrs and some patients will start their session in the afternoon. In these patients
continuation of monitoring for 24hrs will be required to capture the “normal waking hours”
across a spread of 24hrs. Consideration would also need to be given to shift and night workers
whose “normal waking hours” will differ.
When ABPM is poorly tolerated, inconvenient for the patient, or the patient does not want to
undergo ABPM, HBPM should be offered to establish the diagnosis of hypertension. HBPM
may also be preferred to monitor the control of blood pressure in treated patients with a
significant white coat effect, or where this is the patient’s preference for monitoring their
blood pressure control (see section 8.2 – monitoring blood pressure control). Regarding use
of HBPM, the GDG noted that a range of strategies had been used in studies to establish the
HBPM average blood pressure reading. The optimal timing of measurements and the number
96
Hypertension (partial update)
Diagnosis of Hypertension
of measurements required was reviewed. The GDG concluded that a standardised approach
was needed and recommended that patients should measure their blood pressure whilst seated
and relaxed and that at each measurement session, two blood pressure measurements should
be taken, at least one minute apart, in the morning and the evening. The recording should
continue for at least 4 days and ideally 7 days. The readings on the first day should be
discarded and the readings for all remaining days should be used to establish the HBPM
average.
The GDG discussed a number of caveats to recommendations regarding the use of ABPM to
establish the diagnosis of hypertension. Some people may have severe hypertension at
screening with CBPM (i.e. systolic BP ≥180mmHg and/or diastolic BP ≥110mmHg) and in
such cases, clincians should not delay treatment whilst awaiting the results of ABPM – in
these cases, the subsequent ABPM will serve to confirm the diagnosis and severity of the
hypertension; ii) some people will have atrial fibrillation or other significant pulse irregularity
that might render automated BP monitoring (ABPM and HBPM) inaccurate or impossible, in
such cases manual auscultation of blood pressure in the clinic would be the only alternative;
and iii) some people may not tolerate ABPM – in these people HBPM can be used an
alternative on the grounds of better prognostic value and better specificity for hypertension.
However, the GDG noted that based on current data, HBPM could not be considered
equivalent to ABPM with regard to accuracy of diagnosis and emphasised that that ABPM is
the preferred means of confirming or refuting the diagnosis of hypertension.
The GDG also discussed whether ABPM was necessary for confirmation of diagnosis in all
patients, or whether it could be used more selectively, e.g. only in those close to the
diagnostic threshold. The GDG noted that even in people with stages 2, or resistant
hypertension, a significant white coat effect can occur, which would be important to
document to facilitate decisions about the best strategy for subsequent monitoring of blood
pressure control on treatment. The need for ABPM for people with evidence of target organ
damage, e.g. LVH or albuminuria was also discussed by the GDG. It was noted that target
organ damage may not always be due to hypertension, even when the two appear to co-exist.
For example, the presence of ECG LVH in a patient subsequently shown not to be
hypertensive on ABPM would prompt consideration of alternative causes for the ECG
abnormality. Furthermore, some people have higher blood pressures away from the clinic (so
called masked hypertension) and ABPM could reveal much worse blood pressure control
levels than apparent in the clinic – this would be important to know. Finally, the GDG noted
that people with target organ damage are a higher risk group and the best possible assessment
of their blood pressure level when initiating treatment seemed appropriate, mindful of the
better prognostic value of ABPM when compared to CBPM. Overall, the GDG could not
identify a strong evidence-base or clinical argument against the use of ABPM to improve the
accuracy of diagnosis of hypertension, which for many people results in exposure to life-long
treatment. The residual concern in the GDG deliberations was not whether this was the right
thing to do but rather, whether the strategy would be cost-effective (see below) and whether
the practical challenges of implementing an ABPM-based strategy for diagnosis could be
overcome.
The GDG were also mindful of the concerns about the accuracy of automated devices for
measuring blood pressure in people with atrial fibrillation and considered this an important
area for technology development to see if such problems can be overcome. The GDG noted
that In some patients with chronic atrial fibrillation with good rate control, automated devices
can function effectively but concluded that until automated devices, validated for routine
clinical use are available for people with atrial fibrillation, manual auscultation over the
brachial artery is the only practical alternative to measure blood pressure in people with
significant cardiac rhythm irregularity.
97
Upda
Hypertension (partial update)
Diagnosis of Hypertension
As noted above, evaluation of the effectiveness of different methods for measuring blood
pressure to establish the diagnosis of hypertension suggested that ABPM would be the most
accurate method, avoiding clinical disease labelling and treatment of people who were not
truly hypertensive according to their ABPM average blood pressure. The GDG noted,
however, that despite the clear effectiveness of ABPM in improving the specificity and
sensitivity of diagnosis for hypertension, ABPM devices are considerably more expensive
than simple desk top blood pressure monitors and the GDG recognised the obvious potential
cost implications of recommending the more widespread use of ABPM for the routine
diagnosis of hypertension. The GDG thus identified modelling of the cost effectiveness of
different methods for blood pressure measurement as the highest priority for economic
analysis as a prior literature search had identified no published work addressing this key
question in sufficient detail.
The cost-effectiveness analysis compared CBPM, HBPM or ABPM for confirming a
diagnosis in people with suspected hypertension. The GDG spent considerable time
discussing the various factors that would potentially impact on the costs of using ABPM and
also HPBM as an alternative to current standard practice of using a series of CBPM readings
to confirm the diagnosis of hypertension. These included the number and type of healthcare
appointments required to confirm a diagnosis with each method, the failure rate associated
with ABPM and HBPM and the number of uses of the devices each year. As well as initial
diagnosis costs, the analysis took into account downstream costs including hypertension
treatment, checkups and development of cardiovascular disease. Health benefits were
quantified in terms of QALYs. A summary of the cost-effectiveness analysis is provided in
Section 6.3 with full details available in Appendix J: Cost-effectiveness analysis.
Contrary to what might have been expected and mindful of the higher costs of ABPM
devices, the cost-effectiveness analysis found ABPM to be the most cost effective option for
the diagnosis of hypertension across a range of age groups in both men and women.
Remarkably, in most groups ABPM was found to actually improve health (increased QALYs)
and reduce costs, suggesting that use of ABPM for the diagnosis of hypertension has the
potential to be cost saving for the NHS. The GDG noted that this conclusion was robust to a
wide range of sensitivity analyses including those varying the cost of ABPM, the failure rate
for ABPM, the level of CVD risk and the prevalence of true hypertension in the population.
Unsurprisingly, the conclusion was sensitive to assumptions regarding the accuracy of
diagnosis with each method, e.g. when the other methods (CBPM or HBPM) were assumed to
be as accurate as ABPM – which the effectiveness analysis suggests they are not. The
conclusion was also sensitive to the assumption that people who were not hypertensive but
were treated did not receive benefits from treatment, which they might. On the other hand, the
analysis did not model the impact of unnecessarily treating people who are not hypertensive
and the costs, inconvenience, adverse effects of treatment and impact disease labelling may
have on individual patients incorrectly diagnosed as hypertensive.
The extensive GDG deliberations on the cost effectiveness analysis concluded that the use of
ABPM for the routine diagnosis of hypertension, using a daytime average threshold of
≥135/85mmHg, in people who have previously been identified as potentially hypertensive at a
threshold of ≥140/90mmHg using a CBPM, would be both cost-effective and in almost all
cases, cost saving for the NHS, as well as improving the accuracy of diagnosis for patients.
The GDG thus recommended that ABPM should be implemented for the routine diagnosis of
hypertension in primary care.
The GDG also discussed other important aspects when considering the diagnosis of
hypertension including: i) whether there might be an underlying secondary cause for the
elevated blood pressure that might warrant referral for specialist evaluation; ii) whether the
patient might have accelerated hypertension requiring emergency in-patient care; and iii) the
Update 2011
98
Hypertension (partial update)
Diagnosis of Hypertension
need to assess for the presence of target organ damage and formally assess cardiovascular
disease risk.
The GDG recognised and discussed the considerable challenges for implementation of this
recommendation. Sufficient numbers of validated ABPM devices would need to be procured
and adequately maintained. Staff would need to be trained in their use and the interpretation
of data generated by the ABPM reports. The existing recommendations on use of appropriate
cuff size (see section 5.5) and recognition that automated measurements may be unreliable or
impossible in people with significant pulse irregularity (e.g. atrial fibrillation) (see section
5.7) still apply. Some people will not tolerate ABPM and in others the procedure will fail. The
GDG modelled an anticipated failure rate of 5%, ranging to a more extreme failure rate of
10% in sensitivity analyses in the cost effective analysis and ABPM remained the most cost
effective option for the diagnosis of hypertension. In those unable to tolerate or unwilling to
undergo ABPM, the GDG recommended HBPM as an alternative means of confirming the
diagnosis of hypertension with emphasis that ABPM is the preferred method. For those with
significant pulse irregularity, ABPM and HBPM are likely to be unreliable methods for blood
pressure measurement and a series of CBPM readings via manual auscultation (see section
5.4.1) remains the only suitable option.
Finally, the GDG discussed the practicalities of implementing this strategy for the diagnosis
of hypertension. That implementation of this strategy is a challenge is acknowledged.
Presently, some but not all primary care practices have access to ABPM devices, others do
not. Some practices access ABPM through referral to secondary care. Few practices presently
have sufficient numbers of devices to increase their use as required by this guideline
recommendation. The GDG discussed the fact that models of future care cannot just be based
on what we do now and considered it likely that alternative models of service provision would
emerge, reflecting first and foremost what was best and most convenient for patients and local
demand. The GDG considered it inevitable that the costs of ABPM devices will fall as
demand for their use increases and that different models of ABPM provision will evolve over
time to meet local demand.
6.6 Recommendations
8. When considering a diagnosis of hypertension, measure blood pressure in both arms:
 If the difference in readings between arms is more than 20 mmHg, repeat the
measurements.
 If the difference in readings between arms remains more than 20 mmHg on the second
measurement, measure subsequent blood pressure in the arm with the higher reading.
[new 2011]
9. If blood pressure measured in the clinic is 140/90mmHg or higher:
 Take a second measurement during consultation.
 If the second measurement is substantially different from the first, take a third
measurement.
Record the lower of the last two measurements as the clinic blood pressure. [new 2011]
11. If a person is unable to tolerate ABPM, home blood pressure monitoring (HBPM) is a
suitable alternative to confirm the diagnosis of hypertension. [new 2011]
99
Update 2011
10. If the clinic blood pressure is 140/90 mmHg or higher, offer ambulatory blood pressure
monitoring (ABPM) to confirm the diagnosis of hypertension. [new 2011]
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
12. If the person has severe hypertension, consider starting antihypertensive drug treatment
immediately, without waiting for the results of ABPM or HBPM. [new 2011]
13. While waiting for confirmation of a diagnosis of hypertension, carry out investigations
for target organ damage (such as left ventricular hypertrophy, chronic kidney disease and
hypertensive retinopathy) (see 21) and a formal assessment of cardiovascular risk using a
cardiovascular risk assessment tool (see 20). [new 2011]
14. If hypertension is not diagnosed but there is evidence of target organ damage such as left
ventricular hypertrophy, albuminuria or proteinuria, consider carrying out investigations
for alternative causes of the target organ damage. [new 2011]
15. If hypertension is not diagnosed, measure the person’s clinic blood pressure at least every
5 years subsequently, and consider measuring it more frequently if the person’s clinic
blood pressure is close to 140/90 mmHg. [new 2011]
16. When using ABPM to confirm a diagnosis of hypertension, ensure that at least two
measurements per hour are taken during the person’s usual waking hours (for example,
between 08:00 and 22:00). Use the average value of at least 14 measurements taken during
the person’s usual waking hours to confirm a diagnosis of hypertension. [new 2011]
18. Refer the person to specialist care the same day if they have:
 accelerated hypertension, that is, blood pressure usually higher than 180/110 mmHg
with signs of papilloedema and/or retinal haemorrhage or
 suspected phaeochromocytoma (labile or postural hypotension, headache, palpitations,
pallor and diaphoresis). [2004, amended 2011]
19. Consider the need for specialist investigations in people with signs and symptoms
suggesting a secondary cause of hypertension. [2004, amended 2011]
7
Assessing cardiovascular risk, target organ damage and secondary
causes of hypertension
There are four key objectives in the assessment of a person with suspected hypertension; i) to
confirm whether or not blood pressure is elevated (see section xxx); ii) to document the
presence or absence of blood pressure related target organ damage damage (e.g. left
ventricular hypertrophy, hypertensive retinopathy, increased albumin:creatinine ratio); iii) to
evaluate the person’s cardiovascular risk either due to established cardiovascular disease or
high cardiovascular disease risk states (e.g. diabetes or CKD), or by calculation of their 10
year CVD risk estimate (ref section and NICE guidance), and iv) to consider whether their
may be secondary causes for the hypertension.
100
Update 2011
17. When using HBPM to confirm a diagnosis of hypertension, ensure that:
 for each blood pressure recording, two consecutive measurements are taken, at least 1
minute apart and with the person seated and
 blood pressure is recorded twice daily, ideally in the morning and evening and
 blood pressure recording continues for at least 4 days, ideally for 7 days.
Discard the measurements taken on the first day and use the average value of all the
remaining measurements to confirm a diagnosis of hypertension. [new 2011]
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
The risk of clinical events associated with hypertension is not only determined by the level of
blood pressure but also by; i) the presence of target organ damage; ii) the presence of
established cardiovascular disease (iscahemic heart disease or heart failure, cerebrovascular
disease, peripheral vascular disease) or concomitant disease associated with high
cardiovascular disease risk, e.g. diabetes or CKD; or iii) the calculated cardiovascular risk
(estimated from factors such as age, gender, smoking history, etc.). Therefore, routine
assessment of simple markers of target organ damage, a clinical history and examination to
identify associated cardiovascular disease and when indicated, cardiovascular risk calculation,
all form part of the routine assessment of a patient with suspected or confirmed hypertension.
This assessment will also help clinicians to decide the appropriate blood pressure threshold at
which to consider drug therapy for the treatment of hypertension and whether any additional
therapies to reduce cardiovascular disease risk (e.g. statins and antiplatelet therapy) should
also be offered to the patient.
The clinical history, examination and routine blood and urine tests will also alert the clinician
to possible secondary causes of hypertension, some of which are potentially life threatening
(e.g. phaeochromocytoma), and others which might be amenable to potentially curative
interventions (e.g. Conn’s adenoma, fibromuscular dysplasia).
Hypertension and cardiovascular disease
An analysis of 61 prospective observational studies, involving nearly one million individuals,
explored the relationship between blood pressure level and 12,000 strokes and 34,000
ischaemic heart disease events over an average of 13.2 years follow-up361. Across age bands
from 40 to 89, reduction in usual diastolic blood pressure of 20 mmHg systolic or 10 mmHg
diastolic blood pressure was associated with reductions in death from stroke and ischemic
heart disease of about one half, slightly more in the youngest and slightly less in the oldest.
Findings were similar for men and women, for different types of stroke, and consistent across
the range of blood pressure (down to 115/75 mmHg).
An earlier analysis of nine observational studies, involving 420,000 individuals explored the
relationship between blood pressure level and 843 subsequent strokes and 4,856 coronary
events over an average of 7 years follow-up379. Reductions in usual diastolic blood pressure of
5, 7.5 and 10 mmHg were associated with reductions in stroke of 34%, 46% and 56% and
coronary heart disease of 21%, 29% and 37% respectively. The relationship between blood
pressure and disease was constant over a wide range suggesting there is no clear threshold
below which further reduction in blood pressure becomes unbeneficial or harmful.
The implication of these two studies is that some or all of the predicted benefits, found by
comparing individuals with different usual blood pressure levels, could be obtained by one
patient maintaining a similar reduction.
A systematic review of 14 antihypertensive randomised drug trials (diuretics or beta-blockers
compared with placebo) included 37,000 patients135. A mean reduction in diastolic blood
pressure of 5–6 mmHg over 5 years achieved a relative reduction in stroke of 42% (95% CI:
33–50%) and CHD of 14% (95%CI: 4–22%). The authors concluded that virtually all of the
epidemiologically observed benefit from reduced stroke and over half of the reduction in
coronary heart disease could be achieved by lowering blood pressure.
7.1 Routine clinical investigations
A full cardiovascular assessment should be conducted in patients with persistently raised
blood pressure who do not have established cardiovascular disease. There is no firm evidence
from which to define the exact composition of assessment and recommendations are
consensus-based. Medical history, physical examination, and limited diagnostic testing serve
to identify an individual patient's profile of cardiovascular risk factors including age and
gender, smoking, hyperlipidaemia, diabetes, and family history of cardiovascular disease.
101
7.1.2
7.2 Cardiovascular Risk Assessment
Risk models have been developed (as charts, graphs or computer programmes) to allow
clinicians to predict the likelihood of patients developing coronary or cardiovascular disease
using lifestyle and clinical markers (See NICE Lipids Modification, CG67). Although they
vary in detail, risk models may estimate an individual's risk of coronary heart disease and
stroke over the next ten years using their gender, age, diabetic status, smoking status, total
serum cholesterol (TC), high density lipoprotein cholesterol (HDL-C) and blood pressure. An
important aspect of risk models is that they lead the clinician to address a patient's overall
profile of risk rather than treat one risk factor in isolation. Risk factors have a cumulative
effect, and an individual with a number of modest risk factors may be at greater risk of
developing cardiovascular disease than an individual with one high risk factor23. Since several
risk factors are potentially modifiable, an important aspect is which of these to address and in
what order.
7.3 Secondary Hypertension
 An identifiable cause of hypertension is more likely when hypertension occurs in younger
patients (less than 40 years of age), worsens suddenly, presents as accelerated
102
Update 2011
7.1.1
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
Testing may detect diabetes and identify signs of developing target organ damage such as left
ventricular hypertrophy and angina. It may also detect secondary causes of hypertension.
The guideline group identified the following tests as necessary to obtain an accurate profile of
cardiovascular risk. These tests may help identify diabetes, evidence of hypertensive damage
to the heart and kidneys, and secondary causes of hypertension such as kidney disease:
 Urine strip test for blood and protein
 Blood electrolytes and creatinine, and eGFR
 Blood glucose
 Serum total and HDL cholesterol
 12 lead electrocardiogram.
Urine testing for proteinuria
The presence of protein in urine identifies patients with kidney damage, but does not
distinguish between patients who have renal disease and secondary hypertension and those in
whom kidney damage is due to essential hypertension. The test consists of dipping a test strip,
which is impregnated with chemicals which react to protein, into a sample pot of urine. After
30–60 seconds (or according to manufacturer's instructions) the strip is read alongside a
colour code provided. A more sensitive test for urine protein is available by requesting the
local chemical biochemistry laboratory to assay microalbumin in a random specimen of urine.
For further information refer to NICE Clinical Guideline 73.
Blood electrolyte, urea, creatinine, glucose and total/HDL cholesterol levels
These are measured in serum or plasma (glucose) using standard clinical biochemistry
methods. Sodium and potassium levels are checked to exclude hypertension resulting from
adrenal disease. Likewise, urea and creatinine measurements, which reflect kidney function,
are measured to exclude kidney disease as a secondary cause of hypertension Glucose levels
are tested to evaluate diabetes and cholesterol profiles are used to assess cardiovascular risk.
12 lead electrocardiogram. Refer to NICE guidance on Diabetes (Clinical Guidelines 15 and
87).
From an ECG it is possible to determine heart rate, rhythm, conduction abnormalities, left
ventricular size and damage to specific regions of the heart muscle. The presence of
electrocardiographic left ventricular hypertrophy is a variable used in cardiovascular risk
calculators. An echocardiogram might be considered, to confirm or refute the presence of
LVH suggested by ECG findings.
7.3.1
7.3.2
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
hypertension (BP more than 180/110 mmHg with signs of papilloedema and/or retinal
haemorrhage) or responds poorly to treatment. [III]
 An elevated creatinine or reduced eGFR indicates renal disease. Labile or postural
hypotension, headache, palpitations, pallor and diaphoresis are potential signs of
pheochromocytoma. Hypokalaemia, abdominal or flank bruits, or a significant rise in
serum creatinine when starting an ACEi or ARB may indicate renovascular hypertension.
Isolated hypokalaemia may be due to hyperaldosteronism. Potential signs of Cushing
syndrome include osteoporosis, truncal obesity, moon face, purple striae, muscle
weakness, easy bruising, hirsutism, hyperglycemia, hypokalaemia, and hyperlipidaemia.
[III]
Secondary hypertension refers to high blood pressure from an identifiable underlying cause. It
may occur in up to 10% of hypertension cases, the most common cause being chronic renal
disease. Other principal identifiable causes are renovascular hypertension,
pheochromocytoma, Cushing syndrome, and primary aldosteronism. Signs and symptoms of
the main causes of secondary hypertension and available diagnostic tests are summarised
below, although many of these techniques are not provided in primary care but accessed
through specialist referral. We retrieved no useful diagnostic studies which might establish
primary care screening characteristics for secondary causes of hypertension as a basis for
referral: current advice is simply to be aware of signs and symptoms and refer on the basis of
a high index of suspicion and where the findings are likely to necessitate specialist
management.
Renal and renovascular disease
Chronic kidney disease is the most common identifiable cause of hypertension occurring in
2% to 5% of patients182. The British National Formulary advises against routinely using ACEi
or ARBs in patients with known or suspected renovascular disease26.
Signs and symptoms indicating that hypertension may be associated with renal disease are:
young onset of hypertension (before 40 years of age), sudden onset of hypertension or
progressive deterioration in middle age, accelerated hypertension (BP more than 180/110
mmHg with signs of papilloedema and/or retinal haemorrhage), oliguria (urine output <250
ml/day) or anuria (<50 ml/day), oedema, acidosis (acidic blood, <pH), abnormal serum urea
or reduced eGFR, systolic or diastolic bruit467, drug resistant hypertension or increased
creatinine with ACEi or ARB, hypertension onset > 60 years, DBP >110 mmHg, and anaemia
(lowered red blood cell count) resulting in insufficient oxygen to tissues and organs. Although
renal artery stenosis is suggested by the presence of an abdominal or flank bruit, it is an
insensitive test (sensitivity=65%; specificity=90%). When present it is a good marker
(positive likelihood ratio=6.5) but when absent does not rule out renal artery stenosis
(negative likelihood ratio=0.4)182,505.
Renal disease may be diagnosed by elevated serum levels of urea or creatinine (found by a
blood test) or reduced eGFR . Specialist investigation includes magnetic resonance
angiography for imaging of the kidneys, and duplex ultrasound scanning directly measuring
the size of the kidneys467, 35. Test sensitivities have been reported for these investigations182.
Pheochromocytoma
A pheochromocytoma is a tumour which produces and releases large amounts of adrenaline
and noradrenaline (hormones) into the blood. It is rare and may occur in between 0.04% and
0.1% of patients; about 10% are malignant. Adrenaline causes an increase in heart rate and
contractility, while noradrenaline increases systemic vascular resistance. Patients with signs
and symptoms of pheochromocytoma need immediate specialist investigation given the
seriousness of the condition and risk to the patient. The definitive treatment of
pheochromocytoma is surgical removal of the tumour.
103
7.3.3
7.3.4
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
Signs and symptoms include a rapid heart rate, headache, high blood glucose levels, elevated
basal metabolic rate, facial flushing, nervousness, sweating, decreased gastrointestinal
movements and oedema.
Diagnostic techniques include plasma or 24 hour urine collections for metadrenaline and
normetadrenaline 22,250. Following positive findings two types of imaging study may be used
to locate the tumour: metaiodobenzyl-guanidine (MIBG) scintigraphy and computed
tomography (CT).
Hyperaldosteronism (primary aldosteronism)
Aldosterone is a hormone that regulates sodium and water balance. Hyperaldosteronism can
due to bilateral adrenal hyperplasia or Conn’s adenoma occurring in 0.01% to 0.03% of
patients182,570], although its prevalence is contested and may be much higher [364.
Signs and symptoms include sodium retention, and hypokaelaemia leading to heart rhythm
irregularities and possibly muscle weakness. The hypokaelaemia may only occur when
diuretic-induced hypokalaemia is not explained by natural causes467.
Measurement of plasma aldosterone levels and plasma renin activity as the aldosterone:renin
ratio may be used to detect primary aldosteronism250. As with any laboratory test,
standardisation of laboratory assays is important.
Cushing's syndrome
Cushing's syndrome is a syndrome generated by excess glucocorticoids. Cushing’s Disease
specifically refers to over-production of ACTH by the pituitary gland and is the most common
form of the syndrome. Over-production of cortisol can also be due to a tumour in the adrenal
gland, either benign (an adenoma), or malignant (a carcinoma) and in this variant is not
dependent on ACTH. Production of ACTH in an organ or gland other than the pituitary or
adrenal gland (e.g. thymus gland, lung, pancreas) is called ectopic corticotrophin-releasing
production469. Cushing's syndrome may occur in 0.1% to 0.6% of patients.
Signs and symptoms include hypertension, sudden onset of weight gain, central obesity, moon
face, weakness, fatigue, backache, headache, glucose intolerance, oligomenorrhoea
(infrequent menstruation), amenorrhoea (abnormal discontinuation of periods), increased
thirst, increased urination, impotence, muscle atrophy, depression, insomnia, thinning of the
skin, cutaneous hyperpigmentation (darkening of the skin), osteoporosis469.
Diagnosis of Cushing's syndrome begins with a single dose overnight dexamethasonesuppression test. A differential diagnosis is achieved by measuring plasma ACTH together
with either a long dexamethasone suppression test or a corticotrophin-releasing hormone
(CRH) stimulation test217,437.
7.4 Other identifiable causes of hypertension
7.4.1
7.4.2
Hypothyroidism
Hypothyroidism is under production of the hormone thyroxine (which controls metabolism)
by the thyroid gland. Hypertension in hypothyroid patients may result from altered levels of
renin, angiotensin and aldosterone. After thyroid replacement therapy diastolic blood pressure
returns to normal in patients with hypothyroidism suggesting a cause-and-effect
relationship185,329,509. Signs and symptoms include lethargy, fatigue, weight loss, hair loss,
confusion, nausea, bone pain, muscle weakness, slow heart rate. Hypothyroidism is associated
with increased diastolic blood pressure75,572. Hypothyroidism is diagnosed by measuring
thyroid stimulating hormone levels467.
Hyperthyroidism
Hyperthyroidism is the excessive secretion of thyroxine by the thyroid gland. Signs and
symptoms include increased systolic blood pressure, increased metabolic rate, enlargement of
the thyroid gland, tachycardia (increased heart rate), exophthalmia (abnormal protrusion of
the eyeball in the orbit), oedema, dry hair and skin, weight gain, goitre (enlarged thyroid
gland)314. Hyperthyroidism is diagnosed by measuring thyroid stimulating hormone levels467.
104
7.4.3
7.4.4
7.4.5
Recommendations
For NICE guidance on the early identification and management of chronic kidney disease see
‘Chronic kidney disease’ (NICE clinical guideline 73, 2008).
20. Use a formal estimation of cardiovascular risk to discuss prognosis and healthcare
options with people with hypertension, both for raised blood pressure and other modifiable
risk factors. [2004]
21. Estimate cardiovascular risk in line with the recommendations on Identification and
assessment of CVD risk in ‘Lipid modification’ (NICE clinical guideline 67)c. [2008]
22. For all people with hypertension offer to:
 test for the presence of protein in the urine by sending a urine sample for estimation of
the albumin:creatinine ratio and test for haematuria using a reagent strip
 take a blood sample to measure plasma glucose, electrolytes, creatinine, estimated
glomerular filtration rate, serum total cholesterol and HDL cholesterol
 examine the fundi for the presence of hypertensive retinopathy
 arrange for a 12-lead electrocardiograph to be performed. [2004, amended 2011]
c
Clinic blood pressure measurements must be used in the calculation of cardiovascular risk.
105
Update 2011
7.4.6
Hypertension (partial update)
Assessing cardiovascular risk, target organ damage and secondary causes of
hypertension
Obstructive sleep apnoea
Obstructive sleep apnoea is caused by the upper airway becoming obstructed during sleep. It
is more prevalent in men. Signs and symptoms include daytime somnolence (unnatural
drowsiness and sleepiness), obesity, snoring, lower extremity oedema, nocturia and morning
headaches. The main diagnostic technique is a polysomnograph to monitor normal and
abnormal physiological activity during sleep 250,467. Please refer to NICE Technology
Appraisal 139 (www. http://guidance.nice.org.uk/TA139/Guidance/pdf/English) for guidance
on continuous positive airway pressure (CPAP).
Coarctation of aorta
Coarctation of aorta is a congenital condition where a segment of the aorta is too narrow,
reducing oxygenated blood flow around the body. Signs and symptoms include high blood
pressure, decreased or delayed femoral pulse, abnormal chest radiograph. Diagnostic
techniques: doppler or CT imaging of the aorta467.
Acromegaly
Acromegaly is due to excess production of growth hormone. Signs and symptoms of
acromegaly include hypertension, cardiomegaly, enlarged facial features, enlarged jaw,
headache and arthralgia, hypertrichosis, excessive sweating, tiredness, weakness, somnolence
and impaired glucose tolerance360. Acromegaly is diagnosed by evidence of increased growth
hormone secretion360.
Drugs
A number of medications are known to cause raised blood pressure. These include
decongestant found in inhaled cold remedies, may raise diastolic blood pressure517,547. Oral
contraceptive pills containing oestrogen may cause small, and occasionally pronounced, rises
in blood pressure. In rare cases accelerated hypertension may occur535. Other drugs that may
raise blood pressure include immunosuppressive agents, nonsteroidal anti-inflammatory
drugs, COX-2 inhibitors, weight loss agents, stimulants (for example, cocaine),
mineralocorticoids, antiparkinsonian agents, monoamine oxidase inhibitors, anabolic steroids,
sympathomimetics467.
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
Research recommendations
2. In people aged under 40 with hypertension, what is the most accurate method of assessing
the lifetime risk of cardiovascular events and the impact of therapeutic intervention on this
risk?
Current short-term (over 10 years) risk estimates are likely to substantially underestimate the
lifetime cardiovascular risk of younger people (aged under 40) with hypertension, because
short-term risk assessment is powerfully influenced by age. Nevertheless, the lifetime risk
associated with untreated stage 1 hypertension in this age group could be substantial. Lifetime
risk assessments may be a better way to inform treatment decisions and evaluate the cost
effectiveness of earlier intervention with pharmacological therapy.
8
Initiating and monitoring treatment, including blood pressure
targets
Blood pressure thresholds for initiating pharmacological treatment
Review question: In adults with primary hypertension, at what blood pressure should
treatment be initiated?
Clinical evidence
The literature was searched for studies published since the original guideline (2003 onwards).
All study types were included, if the population did not consist of people who were
106
Update 2011
The diagnostic threshold for defining hypertension has been progressively lowered over the
past 50 years as treatment of hypertension has been shown to be beneficial at reducing
cardiovascular morbidity and mortality when initiated at progressively lower blood pressure
thresholds. During that time, the focus also shifted from hypertension diagnosed purely on the
basis of diastolic pressure towards systolic pressure thresholds being the most common
indication for treatment – this reflects the increased prevalence of hypertension with ageing
and the usual progressive rise in systolic pressure with age. In the 2004 guideline, two
different grades of hypertension were defined, Grade 1 hypertension (140-159/90-99mmHg)
and Grade 2 hypertension (i.e ≥160/100mmHg).
The guideline recommended that patients with Grade 2 hypertension should be offered
pharmacological treatment. The guideline was more cautious with regard to pharmacological
treatment for uncomplicated Grade 1 hypertension (i.e. in those without evidence of target
organ damage, cardiovascular disease, CKD or diabetes or at a calculated 10 year CVD risk
<20%). This 2011 guideline partial update reviewed evidence published since the cut point of
the last review (2003) to determine whether the existing recommendations for blood pressure
thresholds for diagnosis and treatment of hypertension should be revised. Furthermore, in
light of the recommendation in this guideline update that an ABPM daytime average blood
pressure will hereafter be the preferred method for confirming the diagnosis of hypertension,
the thresholds for diagnosis and grades of hypertension also needed to be reviewed with
regard to ABPM daytime averages.
Once a decision has been made to initiate pharmacological treatment for hypertension, the
next key question was “how low should blood pressure be lowered?” i.e. what is the
recommended blood pressure target? The 2004 guideline noted that the evidence base to
support a recommendation for an optimal treatment target for hypertensiion was less
substantial than it should be. International consensus has specified an optimal treatment
target for hypertension of <140/90 mmHg and in some cases even lower targets for people
with established cardiovascular or renal disease or diabetes. There has also been concern but
little evidence, as to the efficacy, safety and appropriate blood pressure target for the people at
advanced age with hypertension (greater than 80 years). Consequently, studies examining
optimal treatment targets have been reviewed.
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
107
Update 2011
exclusively diabetic or had CKD. Studies were excluded if they did not stratify results into
more than one different BP value / threshold.
Thirty studies (31
papers)49,50,54,57,60,61,68,89,101,119,136,165,206,208,213,243,244,247,269,285,291,313,331,332,340,351,454,466,521,546,629
were found that fulfilled the inclusion criteria and assessed at what BP should treatment be
initiated (appropriate threshold for intervention). One of the studies60,61 was published as two
separate papers reporting different assessment outcomes, so this study has only been counted
once, however results from both papers are reported and referenced here.
The studies addressing the question were categorised into three different types:
1. SRs / MAs (three studies)54,206,351 . The SRs/MAs were of high quality however the studies
they included were either low quality (observational)54,206 or low to high (RCTs).351.
2. Prognostic studies (27 studies; 28
papers)49,50,57,60,61,68,89,101,119,136,165,208,213,243,244,247,285,291,313,331,332,340,454,466,521,546,629 - those that
assess the risk of developing clinical outcomes (over time) at different BP values. Most of the
prognostic studies were found to be methodologically sound (see quality assessment summary
tables in appendix F) except for the following eight studies which had (or were rated as
‘unclear’ for) three or more of the six potential methodological flaws (Fagard 2007,
Gudmundsson 2005, Obara 2007, Okayama 2006, Sleight 2009, Fagard 2004, Britton 2009,
Conen 2007101,136,206,208,243,454,466,546).
Prognostic studies were divided into four categories: those that assessed BP measured by
either clinic, home, ambulatory or self-reported / unknown methods.
3. Blood pressure equivalence studies (one study)269– those that calculate equivalent blood
pressures using different measurement methods (home, ABPM or clinic), in order to set
thresholds for the diagnosis and treatment of HT. All these studies were observational and
therefore low quality.
Data from the included studies was not pooled into a meta-analysis. This was because for
many studies only HRs were given rather than the number of patients with events, and data
was often stratified differently in the studies (for example, by age, gender, treated/untreated
or other population characteristics), making it not possible to pool together. Additionally, it
was deemed inappropriate to pool the studies because the studies themselves differed
considerably in their design and analysis, particularly regarding the following areas:
 blood pressure values, groups and thresholds used
 blood pressure measurement methods used
 outcome measures (and definitions of outcomes) used
 follow-up times used
 covariates taken into account in analyses
Details of all the studies are included in Table 27, Table 28 and Table 30. Table 29
summarises the numerical results for selected outcomes of the prognostic studies included for
this review. The full data for all outcomes can be found in the evidence tables in the appendix.
Systematic reviews/Meta-analyses
N
Population
Asayama et
al., 200954
4571
General
population
(HT and
NT)
MA of data
from 4 cohort
studies
1
SR/MA of 108
RCTs
248,445
HT and NT
0
8
People of
any age,
disease
status, preTreatment
BP and use
of other
drugs
Clinic
Mean
3.5
years
Study design
Outcomes
Prognostic:
Risk (HR) of
developing
clinical
outcomes
Stroke;
death from
stroke
BP difference
trials designed
to achieve a
difference in
BP between
randomised
groups
CHD
events;
stroke
BP values at baseline (groups
/ thresholds); mmHg
Optimal: <120/ <80
Normal: 120-129/80-84
High normal: 130-139/85-89
Grade 1 (mild) HT: 140-159/
90-99
Grade 2 (moderate) HT: 160179/ 100-109
Grade 3 (severe) HT: ≥180/110
Best BP threshold (authors’
conclusions)
Untreated groups: risk (HR) of
first stroke increased linearly
with BP.
10mm SBP increments from
120 – 180 mmHg
BP treatment reduced risk of
CVD and stroke, regardless of
patients’ pre-treatment BP (as
low as 110 SBP and 70 DBP;
mmHg).
Treated people with optimal
BP had higher risk of stroke
than untreated people with
optimal BP.
Lowering BP by 10mmHg SBP
or 5mmHg DBP reduced CVD
events by around 25%, heart
failure (by about 25%) and
stroke (by about 33%).
Authors concluded that BP
lowering drugs should be
offered to anyone at high risk
(whatever the reason for high
risk, e.g. age, cardiovascular
disease event) not just to people
with high BP, because a given
BP reduction lowers the risk of
Update 2011
Law et al.,
2009351
Followup
Mean
9.5
years
Hypertension (partial update)
Reference
BP
measureme
nt method
Clinic
Initiating and monitoring treatment, including blood pressure targets
Table 27: Study details and results for SRs/MAs assessing the risk of developing clinical outcomes at different BP thresholds.
Population
Fagard et al.,
2007206
11,502
General
population,
primary
care and
secondary
care
(HT and
NT)
SR/MA of 7
studies
Clinic and
ABPM (to
give
diagnoses)
Followup
Mean 8
years
Study design
Outcomes
Risk of
developing
events in
people
diagnosed as
NT, WCH,
MH or
sustained HT
CV events
BP values at baseline (groups
/ thresholds); mmHg
Best BP threshold (authors’
conclusions)
coronary heart disease and
stroke by a constant proportion
irrespective of pre-treatment
BP.
NT: normal BP clinic and
ABPM; mean BP 121.8/75.6
and 119.7/72.6 respectively
NS difference between WCH
and NT for incidence of CV
events;
worse CV events in MH and
sustained HT
WCH: clinic HT, normal
ABPM; mean BP 148.2/86.2
and 125.6/74.9 respectively
MH: normal clinic, ABPM HT;
mean BP 129.9/78.6 and
141.1/83.2 respectively
1
Sustained HT: clinic HT and
ABPM HT; mean BP
157.7/88.5 and 152.4/85.7
0
9
HT diagnosis - cut off BP
Clinic: 140/90 mmHg
ABPM: 135/85 mmHg (except
1 study 135/83mmHg)
Hypertension (partial update)
N
Initiating and monitoring treatment, including blood pressure targets
Reference
BP
measureme
nt method
Prognostic studies
N
BP values at baseline (groups /
thresholds); mmHg
Best BP threshold (authors’
conclusions)
Stroke, CV
events
SBP values
<120 (median 114)
120-139 (median 130)
140-159 (median 149)
≥160 (median 169)
The benefits of treatment were
comparable for patients who
were or were not HT at baseline,
for baseline BP levels extending
down to 115/75mmHg.
Risk of developing
events in people
with different
baseline BP values
(grouped)
Stroke
Optimal: <120 /<80
Normal: 120-129 /80-84
High normal: 130-139 /85-89
Grade 1 HT: 140-159 /90-99
Grade 2 HT: 160-179 /100-109
Grade 3 HT: ≥180 /110
Age-adjusted incidence of total
stroke rose progressively with
higher BP in both genders
10 years
Risk of developing
events in people
with different
baseline BP values
(grouped)
Major
coronary
event
NT: ≤140 /90
New HT: SBP >159 and/or
DBP>94
Adequately treated HT: <160 /95
Inadequately treated HT:
≥160/95
In all HT men, including those
receiving “adequate”
antihypertensive Tx, the 10-year
risk of CHD was at least
doubled.
Median 20
years
Risk of developing
events in people
with different
baseline BP values
(grouped)
Study 1:
Mortality
(all cause
and CV)
NT:<160/95 and no Tx
HT (≥160 SBP or 95 DBP or Tx
in last 7 days); treated and
controlled (<160/95mmHg)
HT: Tx and not controlled
HT and aware (HT diagnosis or
current Tx) but untreated
HT but unaware
In men, all-cause and
cardiovascular mortality were
significantly higher in all
hypertensive groups compared
with the normotensive group. In
women, the mortality in those
whose hypertension was
controlled was not significantly
different from the normotensive
group, suggesting that these
Population
Follow-up
Study design
Outcomes
6105
HT and NT
(Cerebrovas
cular
disease)
Mean 3.9 years
Risk of developing
events in people
with different
baseline BP values
1621
General
population
(HT and
NT)
32 years
5389
General
population
(HT and
NT)
Barengo et al.,
2009 and
200960,61
41,895
(study
1)
General
population
(HT and
NT)
Cohort
47,610
(study
2)
Clinic BP measurements
Arima et al.,
200649
Sub-analysis of
RCT
(PROGRESS)
Arima et al.,
200950
Cohort
(HISAYAMA)
1
Assmann et al.,
200557
1
0
Cohort
(PROCAM)
Study 2:
stroke (fatal
or nonfatal)
Hypertension (partial update)
Reference
Initiating and monitoring treatment, including blood pressure targets
Table 28: Study details and results for prognostic studies assessing the risk of developing clinical outcomes at different BP thresholds
Reference
N
Population
Follow-up
Study design
Outcomes
BP values at baseline (groups /
thresholds); mmHg
1
Carlsson et al.,
2009119
1
1
2280
General
population
(HT and
NT)
26 years
Risk of developing
events in people
with different
baseline BP values
(grouped)
Mortality;
CV
mortality
NT/optimal: <130 / <85
Pre-HT: 130-139 and/or 85- 89
DBP
High: 140 - 159 and/or 90-94
DBP
Very high: ≥160 and/or DBP
≥95
Risk of Events increased with
increasing BP; Very high blood
pressure (≥160/95mmHg) is an
independent risk factor for allcause and CV mortality in men
and women.
3246
General
population
(HT and
NT)
Up to 20 years
(mean 13.6 for
men and 14.4
for women)
Risk of developing
events in people
with different
baseline BP values
(grouped)
Mortality;
CV
mortality
NT/high-NT:<140 /<90
Mild-moderate HT: 140-179 /90109
Severe HT: ≥180 /≥110
Patients treated for HT whose
BP is not controlled have a
higher risk of mortality than
those whose BP is controlled.
Cohort study
Gudmundsson
et al., 2005243
Cohort study
(Note: Tx target
<160/<95mmHg; treatment not
as aggressive as it would be
today; number controlled to
<140/90mmHg was less than
Hypertension (partial update)
The risk of stroke was
significantly higher in men and
women in all hypertensive
groups compared with the
normotensive group. It may be
higher in treated than untreated
patients if they have had
hypertension longer and it is
more severe (also unaware were
significantly younger so had
lower risk).
Initiating and monitoring treatment, including blood pressure targets
Best BP threshold (authors’
conclusions)
women benefitted from
achieving normal BP, although
the uncontrolled, untreated and
unaware groups had higher
mortality.
Population
Follow-up
Study design
Outcomes
Ishikawa et al.,
2008291
11,103
General
population
(HT and
NT)
Mean 10.7
years
Risk of developing
events in people
with different
baseline BP values
(grouped)
Stroke
Cohort (JMS)
Kagiyama et
al., 2008313
BP values at baseline (groups /
thresholds); mmHg
Best BP threshold (authors’
conclusions)
half those labelled “controlled”
in this study.)
NT: <140/90, no treatment
HT: treated (receiving Tx,
irrespective of current BP)
C: Controlled (<140/90)
U: Uncontrolled (≥140 and/or
DBP ≥90)
HT: untreated (≥140 /90 without
Tx)
M: Mild (SBP 140-159 or DBP
90-99)
MS: Moderate-severe (SBP ≥160
and/or DBP ≥100)
Risk of stroke higher among HT
vs. NT patients, and treated vs.
non-treated HT, even when BP
controlled to <140/90mmHg
Untreated HT might have had a
shorter duration of HT (and
therefore lower risk of stroke) or
have WCH (also lower risk).
1
639
General
population
(HT and
NT) but
elderly (80
years)
4 years
Risk of developing
events in people
with different
baseline BP values
(grouped)
Mortality
and CV
mortality
SBP values
NT: <140
Mild HT: 140-159
moderate-severe HT: >160
No association between total
mortality and SBP in the very
elderly overall (however
increased risk with increase BP),
but there was an association in
those with CVD or on Tx.
5494
General
population
(HT and
NT)
Mean 11.7
Risk of developing
events in people
with different
baseline BP values
(grouped)
CV events
(MI or
Stroke)
Optimal: <120 /<80
Normal: 120-129 /80-84
High normal: 130-139 /85-89
Stage 1 HT: 140-159 /90-99
Stage 2/3 HT: ≥160 /≥100
Normal and high normal BP
were a risk factor for the
incidence of stroke and MI in
men compared with optimal BP,
as well as hypertension stage 1
or more. In women, the risk was
seen at hypertension stages but
not at normal/high normal BP
(although numbers of events
were lower in women).
Cohort
1
2
Kokubo et al.,
2008331
Cohort
(SUITA)
Very few people in stage 3 so
combined into ‘stage 2’ values
Kono et al.,
2005332
708
HT (with
vs. without
CV event)
n/a as casecontrol study
Risk of developing
events in people
with different
baseline BP values
CV events
SBP values
NT: <140
Mild HT: 140-159
Positive relationship between BP
status and risk of cardiovascular
events
Hypertension (partial update)
N
Initiating and monitoring treatment, including blood pressure targets
Reference
CVD
Optimal: <120 /<80
Normal: 120-129 /80-84
High normal: 130-139 /85-89
Normal BP and high normal BP
were associated with a greater
risk of incident cardiovascular
disease compared with optimal
BP. The risk was also higher for
black people of African and
Caribbean descent, older people
(55-64 compared with 45-54),
those with diabetes, high BMI,
raised LDL cholesterol or renal
insufficiency.
Risk of developing
events in people
with different
baseline BP values
(grouped)
Onset of or
death due to
circulatory
disease
(stroke,
angina, MI,
cardiac
death)
Optimal: <120 /<80
Normal: 120-129 /80-84
High normal: 130-139 /85-89
Grade 1 HT: 140-159 /90-99
Grade 2 HT: 160-179 /100-109
Grade 3 HT: ≥180 /110
In a relatively old cohort (mean
age 60 years), risk of
cardiovascular disease increased
in higher BP groups
Risk of developing
events in people
with different
baseline BP values
(grouped)
Mortality;
CV
mortality
SBP values
Group 1: <120
Group 2: 120-139
Group 3: 140-159
Group 4: 160-179
Group 5: >179
Increased BP associated with
cardiovascular disease mortality
at all ages
Population
Follow-up
Study design
(grouped)
Outcomes
Kshirsagar et
al., 2006340
8960
General
population
(HT and
NT)
Mean 11.6
years
Risk of developing
events in people
with different
baseline BP values
(grouped)
10,300 personyears
19 years
Cohort (ARIC)
Obara et al.,
2007454
1798
General
population
(HT and
NT)
Post-hoc
analysis
(cohort)
1
1
3
Okayama et
al., 2006466
Cohort
(NIPPON
DATA 80)
4244
General
population
(HT and
NT)
DBP values
Group 1: <80
Group 2: 80-84
Group 3: 85-89
Group 4: 90-99
Group 5: >99
Hypertension (partial update)
Best BP threshold (authors’
conclusions)
N
Initiating and monitoring treatment, including blood pressure targets
BP values at baseline (groups /
thresholds); mmHg
moderate-severe HT: >160
Reference
Case-control
Population
Follow-up
Study design
Outcomes
Sairenchi et
al., 2005521
97,153
General
population
(HT and
NT)
Mean 8.7 years
(men), 8.9
years (women)
Risk of developing
events in people
with different
baseline BP values
(grouped)
Mortality
25,558
People with
atherosclero
tic disease
or diabetes
with end
organ
damage
(High risk)
Mean 56
months
Risk of developing
events in people
classed into
baseline BP
quartiles
CV events
(CV death,
MI, Stroke,
HF)
Cohort
Sleight et al.,
2009546
Post-hoc
analysis of
RCT
(ONTARGET)
BP values at baseline (groups /
thresholds); mmHg
Optimal: <120 /<80
Normal: 120-129 /80-84
High normal: 130-139 /85-89
Stage 1 HT: 140-159 /90-99
Stage 2/3 HT: ≥160 /≥100
Best BP threshold (authors’
conclusions)
Impact of SBP and DBP on
cardiovascular disease around 2
times larger among middle-aged
than elderly subjects (men and
women); generally an increase in
risk with increase BP values
SBP values (quartiles)
≤130 mmHg
130-142 mmHg
142-154 mmHg
>154 mmHg
No relationship found between
SBP reduction and risk of MI,
congestive heart failure and
cardiovascular death.
Avoid excessive SBP reduction
(below 130mmHg) in older
sicker high-risk patients
1
For the primary outcome, there
is a J-shaped pattern (nadir
130mmHg) in the relationship
between on-treatment SBP
(deciles) and adjusted risk of
events; this was also true for
cardiovascular mortality (nadir
130mmHg) and MI (126mmHg)
but not for stroke.
1
4
Haider et al.,
2003247
Cohort
(Framingham
heart study
subset)
2040
General
population
Mean 17.4
years
Risk of developing
events in people
classed into
baseline BP groups
Congestive
HF
SBP values
87-125 mmHg
126-141 mmHg
≥161 mmHg
DBP values
49-74 mmHg
75-82 mmHg
Both SBP and DBP were
associated with CHF, but SBP
conferred greater risk than DBP.
Increased risk of events with
increased BP value.
Hypertension (partial update)
N
Initiating and monitoring treatment, including blood pressure targets
Reference
Population
Follow-up
Study design
Outcomes
Benetos et al.,
200368
34,776
NT, HT and
HT (Tx)
8-12 years
Risk of developing
events in people
iwth higher and
lower BP values
(and in Tx and unTx HT).
CVD, CHD
and
associated
mortality
Case-control
Best BP threshold (authors’
conclusions)
Treated (mean BP ~151/93
mmHg)
Untreated (mean BP ~136/83
mmHg)
High BP (≥140/90 mmHg)
Lower BP(<140/90)
Treated HTs had higher SBP (+
15 mmHg) and higher DBP (+ 9
mmHg), and a higher prevalence
of associated risk factors and
diseases. Treated HTs vs.
untreated HTs presented a twofold increase in the RR for CV
mortality and CHD mortality.
Adjustment for unmodifiable
risk factors only slightly
decreased the excess CV risk
observed in treated people. After
additional adjustment for
modifiable associated risk
factors, the increased mortality
in treated people persisted. Only
after additional adjustment for
SBP were CV mortality and
CHD mortality similar in the two
groups of people.
1
BP values at baseline (groups /
thresholds); mmHg
≥83 mmHg
1
5
Therefore, the increased CV
mortality in treated HT vs.
untreated HT is mainly due to
high SBP levels under treatment.
Weitzman et
al., 2006629
Cohort
9611
General
population
(HT and
NT)
23 years
Risk of developing
events in people
classed into
baseline BP groups
Mortality
(stroke,
CHD and
all-cause)
SBP values
80-119 mmHg
120-129 mmHg
130-136 mmHg
137-149 mmHg
150-260 mmHg
DBP values
Hypertension (partial update)
N
Initiating and monitoring treatment, including blood pressure targets
Reference
Population
Follow-up
Study design
Outcomes
Borghi et al.,
200389
2939
General
population
(HT and
NT)
23 years
Risk of developing
events in people
classed into
baseline BP groups
Mortality,
CHD, MI,
CeVD
Cohort
(Brisighella
Heart Study)
BP values at baseline (groups /
thresholds); mmHg
40-77 mmHg
78-80 mmHg
81-85 mmHg
86-90 mmHg
91-150 mmHg
Best BP threshold (authors’
conclusions)
SBP values
<120 mmHg
120-139 mmHg
140-159 mmHg
>159 mmHg
There is a consistent, strong,
graded association between SBP
(but not DBP) and
cardiovascular events
DBP values
<70 mmHg
70-79 mmHg
80-89 mmHg
>89 mmHg
1
Fang et al.,
2006213
26,587
1
6
Cohort
General
population
(HT and
NT)
Mean 9.5 years
Risk of developing
events in people
classed into
baseline BP groups
Stroke
Increase in combined SHD and
cerebrovascular disease risk was
already evident with high-normal
SBP
ISH: ≥140 / <90 mmHg
SDH: ≥140 / ≥90mmHg
IDH: <140 / ≥90 mmHg (with or
without a-HT Tx)
MHT: <140 / <90 (and
controlled BP by a-HT Tx)
NT: <140 / <90 (without history
of HT)
Highest risk of stroke in people
with ISH and SDH vs IDH and
MHT.
Normal ABP: <140mmHg
Abnormal ABP: 140-159mmHg
High ABP: ≥160mmHg
Baseline ABP predicts
cardiovascular events. Increased
events with increase in BP
People with SDH are at the
highest risk of stroke and should
be treated more aggressively.
Home BP measurements – no studies (one included in Fagard meta-analysis)
Ambulatory BP measurements
Fagard et al.,
2004208
Cohort subanalysis of
295
HT (SBP)
Median 7.5
years
Risk of developing
events in people
classed as normal,
abnormal or high
BP
CV events
Hypertension (partial update)
N
Initiating and monitoring treatment, including blood pressure targets
Reference
ISH determined by ABPM was
associated with a high risk of
stroke, similar to that found for
patients with combined systolicdiastolic HT.
Risk of developing
events in people
classed as NT,
WCH and HT
Death and
CV events
NT: <140; mean = 129.1 mmHg
HT: SBP >140; mean = 160.3
mmHg
WCH: CBP>140, mean = 136.3;
ABPM <135/90 mmHg
There is an increased
cardiovascular risk in WCH
compared to normotensive
controls; the level of risk is the
same as that seen with EHs
(even though WCH had a lower
average ABP than NT).
Risk of developing
events in people
with different
baseline BP values
HF
SBP values
Linear relationship between NT
SBP (120-129mmHg and 130139mmHg) and risk of heart
failure risk, as well as for HT
SBP
Study design
Outcomes
Inoue et al.,
2007285
1,271
HT
Mean 11.2
years
Risk of developing
events in people
classed as HT
(SBP-DBP; ISH,
IDH) vs. NT
566
General
population
(NT, HT
and WCH)
Mean 10.2
years
Cohort
Self-reported / unknown BP measurement method
Britton et al.,
2009101
18,876
HT
Mean 20.7
years
Cohort
NT (not on Tx)
<120 mmHg
120-129 mmHg
130-139 mmHg
HT (or on Tx)
<130 mmHg
130-139 mmHg
140-149 mmHg
150-159 mmHg
≥160 mmHg
Conen et al.,
39,322
NT and HT
Median 10.2
Risk of developing
CV death,
Optimal: <120/ <75
The CV risk of women with high
Hypertension (partial update)
NT: <135 / <80 mmHg
SDH: ≥135 / ≥80 mmHg
ISH: ≥135 / <80 mmHg
IDH: <135 / ≥80 mmHg
Follow-up
Initiating and monitoring treatment, including blood pressure targets
Stroke
Population
Gustavsen et
al., 2003 244
1
Best BP threshold (authors’
conclusions)
N
Cohort; subanalysis of
RCT
(OHASAMA)
1
7
BP values at baseline (groups /
thresholds); mmHg
Reference
RCT (SystEur)
Population
women
Follow-up
years
Study design
events in people
with different
baseline BP values
Outcomes
stroke or
MI
12,218
HT with
CAD
Median 4.1
years
Risk of developing
events in people
with different
baseline BP values
CV death,
non-fatal
MI
Cohort (subanalysis of
RCT)
Deckers,
2006165
Post-hoc
analysis of
RCT
(EUROPA)
BP values at baseline (groups /
thresholds); mmHg
Normal: 120-129/75-84
High normal: 130-139/85-89
HT: ≥140 /≥90
Best BP threshold (authors’
conclusions)
normal BP is higher than those
with normal BP; there was a
strong and consistent increase in
events down to the optimal BP
category.
SBP values
≤130 mmHg
>130-160 mmHg
>160 mmHg
Higher baseline BP associated
with increased risk.
Hypertension (partial update)
N
1
1
8
Initiating and monitoring treatment, including blood pressure targets
Reference
2007136
Table 29: Summary of numerical results for prognostic studies (for selected outcomes)
Arima et al.,
200649
Stroke
Arima et al.,
200950
Stroke
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
SBP values (%, events/ person years) No HR values given
120 (median 114): 6.8%
120-139 (median 130) : 12.2%
140-159 (median 149): 12.5%
≥160 (median 169): 19.0%
Men Optimal: <120 /<80: Reference
Men Normal: 120-129 /80-84: 1.64 (0.76-3.56) p>0.05
Men High normal: 130-139 /85-89: 1.52 (0.70-3.31) p>0.05
Men Grade 1 HT: 140-159 /90-99: 3.31 (1.73-6.32)p<0.05
Men Grade 2 HT: 160-179 /100-109: 4.22 (2.16-8.25)p<0.05
Men Grade 3 HT: ≥180 /110: 5.75 (2.93-11.30)p<0.05
1
Women Optimal: <120 /<80: Reference
Women Normal: 120-129 /80-84: 1.53 (0.60-3.89)p>0.05
Women High normal: 130-139 /85-89: 2.19 (0.93-5.16)p>0.05
Women Grade 1 HT: 140-159 /90-99: 3.92 (1.84-8.35)p<0.05
Women Grade 2 HT: 160-179 /100-109: 4.89 (2.24-10.67)p<0.05
Women Grade 3 HT: ≥180 /110: 7.51 (3.39-16.64)p<0.05
1
9
Assmann et al.,
200557
Barengo et al.,
2009 and 200960,61
Major coronary
event
CV mortality
(MEN)
NT: ≤140 /90
New HT: SBP >159 and/or DBP>94
Adequately treated HT: <160 /95
Inadequately treated HT: ≥160/95
No HR values given
NT:<160/95 and no Tx : Reference
HT (≥160 SBP or 95 DBP or Tx in last 7 days): No HR given
HT treated and controlled (<160/95mmHg) 2.25 (1.70-2.99)
HT: Tx and not controlled 2.41 (2.01-2.89)
HT and aware (HT diagnosis or current Tx) but untreated 1.92 (1.65-2.23)
HT but unaware 1.49 (1.33-1.68)
Hypertension (partial update)
Outcome
Initiating and monitoring treatment, including blood pressure targets
Study
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
CVD, CHD and
associated
mortality
Treated (mean BP ~151/93 mmHg)
Untreated (mean BP ~136/83 mmHg)
High BP (≥140/90 mmHg)
Lower BP(<140/90)
No HRs given
Borghi et al.,
200389
Mortality
SBP values
<120 mmHg Reference
120-139 mmHg 1.48 (1.04-2.10), p=0.0313
140-159 mmHg 1.92 (1.32-2.80), p=0.0006
>159 mmHg 2.38 (1.61-3.50), p<0.0001
Carlsson et al.,
2009119
CV mortality
Men NT/optimal: <130 / <85 Reference
Men Pre-HT: 130-139 and/or 85- 89 DBP 1.07 (0.58-1.97)
Men High: 140 - 159 and/or 90-94 DBP 1.17 (0.66-2.09)
Men Very high: ≥160 and/or DBP ≥95 3.12 (1.84-5.26)
1
Women NT/optimal: <130 / <85 Reference
Women Pre-HT: 130-139 and/or 85- 89 DBP 1.89 (0.76-4.68)
Women High: 140 - 159 and/or 90-94 DBP 2.34 (1.01-5.45)
Women Very high: ≥160 and/or DBP ≥95 3.84 (1.62-9.12)
2
0
Fang et al., 2006213
Stroke
NT: <140 / <90 (without history of HT) Reference
ISH: ≥140 / <90 mmHg 2.35 (1.91-2.90)
SDH: ≥140 / ≥90mmHg 2.96 (2.49-3.52)
IDH: <140 / ≥90 mmHg (with or without a-HT Tx) 2.16 (1.69-2.76)
MHT: <140 / <90 (and controlled BP by a-HT Tx) 1.33 (0.96-1.84)
Gudmundsson et
al., 2005243
CV mortality
Men NT/high-NT:<140 /<90 Reference
Men Mild-moderate HT: 140-179 /90-109 RR: 1.30 (0.79-2.14)
Men Severe HT: ≥180 /≥110 RR: 1.23 (0.72-2.11)
Women NT/high-NT:<140 /<90 Reference
Women Mild-moderate HT: 140-179 /90-109 RR: 1.56 (0.85-2.86)
Hypertension (partial update)
Outcome
Benetos et al.,
200368
Initiating and monitoring treatment, including blood pressure targets
Study
Study
Outcome
Congestive HF
SBP values
87-125 mmHg Reference
126-141 mmHg 1.48 (0.99-2.21), p=0.06
≥161 mmHg 3.07 (2.10-4.49), p<0.001
Ishikawa et al.,
2008291
Stroke
Men NT: <140/90, no treatment Reference
Men HT: treated (receiving Tx, irrespective of current BP) RR:3.00 (2.00-4.51)
Men C: Controlled (<140/90) RR 2.96 (1.66-5.26)
Men U: Uncontrolled (≥140 and/or DBP ≥90) RR 3.05 (1.92-4.85)
Men HT: untreated (≥140 /90 without Tx) RR 2.56 (1.83-3.57)
Men M: Mild (SBP 140-159 or DBP 90-99) RR 2.34 (1.62-3.37)
Men MS: Moderate-severe (SBP ≥160 and/or DBP ≥100) RR 3.17 (2.02-4.97)
Women NT: <140/90, no treatment Reference
Women HT: treated (receiving Tx, irrespective of current BP) RR 3.34 (2.29-4.87)
Women C: Controlled (<140/90) RR 3.69 (2.20-6.17)
Women U: Uncontrolled (≥140 and/or DBP ≥90) RR 3.16 (2.06-4.85)
Women HT: untreated (≥140 /90 without Tx) RR 1.93 (1.35-2.76)
Women M: Mild (SBP 140-159 or DBP 90-99) RR 1.95 (1.32-2.87)Women MS: Moderate-severe (SBP ≥160 and/or DBP
≥100) RR 1.87 (1.08-3.24)
1
Haider et al.,
2003247
2
1
Only RRs given for above categories (but unclear). No HRs given
Kagiyama et al.,
2008313
CV mortality
SBP values
NT: <140: Reference
Mild HT: 140-159: RR:1.71 (0.56-5.24)
moderate-severe HT: >160: RR: 2.15 (0.51-8.97)
Only RRs given for above categories. No HRs given
Kokubo et al.,
CV events (MI
Men Optimal: <120 /<80 Reference
Hypertension (partial update)
Only RRs given for above categories. However, per 1SD rise in SBP (22.4mmHg for men and 22.5 mmHg for women), HRs
for Cv mortality are: 1.00 (0.87-1.15) for men and 1.34 (1.16-1.55),p<0.001 for women
Initiating and monitoring treatment, including blood pressure targets
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
Women Severe HT: ≥180 /≥110 RR: 2.57 (1.36-4.87)
Women Optimal: <120 /<80 Reference
Women Normal: 120-129 /80-84 1.12 (0.59-2.13)
Women High normal: 130-139 /85-89 1.54 (0.85-2.78)
Women Stage 1 HT: 140-159 /90-99 1.35 (0.75-2.43)
Women Stage 2/3 HT: ≥160 /≥100 2.86 (1.60-5.12)
Overall Optimal: <120 /<80 Reference
Overall Normal: 120-129 /80-84 1.62 (1.08-2.43)
Overall High normal: 130-139 /85-89 2.08 (1.42-3.05)
Overall Stage 1 HT: 140-159 /90-99 2.06 (1.42-2.98)
Overall Stage 2/3 HT: ≥160 /≥100 3.53 (2.43-5.13)
1
CV events
SBP values
NT: <140 reference
Mild HT: 140-159 Adjusted OR: 1.69 (1.10-2.60)
moderate-severe HT: >160 Adjusted OR: 2.20 (1.08-4.45)
Only adjusted ORs given. No HRs given
Kshirsagar et al.,
2006340
CVD
Optimal: <120 /<80 Reference
Normal: 120-129 /80-84 1.69 (1.37-2.09)
High normal: 130-139 /85-89 2.33 (1.85-2.92)
Obara et al.,
2007454
Onset of or
death due to
circulatory
disease (stroke,
angina, MI,
cardiac death)
Optimal: <120 /<80
Normal: 120-129 /80-84 Reference
High normal:130-139 /85-89 RR:1.19 (0.89-1.20), p=0.3
Grade 1-3 HT: 140->180 RR: 1.46 (1.00-1.17), p=0.011
Only adjusted RRs given. No HRs given
2
2
Kono et al.,
2005332
Hypertension (partial update)
Outcome
or Stroke)
Initiating and monitoring treatment, including blood pressure targets
Study
2008331
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
Men Normal: 120-129 /80-84 2.04 (1.19-3.48)
Men High normal: 130-139 /85-89 2.46 (1.46-4.14)
Men Stage 1 HT: 140-159 /90-99 2.62 (1.59-4.32)
Men Stage 2/3 HT: ≥160 /≥100 3.95 (2.37-6.58)
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
CV mortality
SBP values
Group 1: <120 Reference
Group 2: 120-139 Age adjusted RR: 2.36 (1.17-4.77)
Group 3: 140-159 Age adjusted RR: 3.00 (1.51-5.94)
Group 4: 160-179 Age adjusted RR: 3.46 (1.75-6.84)
Group 5: >179 Age adjusted RR: 5.13 (2.59-10.16)
No HRs given for categories above, but multivariate adjusted HRs for 1SD increase in SBP: 1.31 (1.17-1.47)
Sairenchi et al.,
2005521
Mortality
Men Optimal: <120 /<80 Reference
Men Normal: 120-129 /80-84 RR: 1.48 (0.50-4.44)
Men High normal: 130-139 /85-89 RR:2.89 (1.07-7.86)
Men Stage 1 HT: 140-159 /90-99 RR:3.06 (1.15-8.16)
Men Stage 2/3 HT: ≥160 /≥100 RR:5.99 (2.13-16.8)
1
Women Optimal: <120 /<80 Reference
Women Normal: 120-129 /80-84 RR:0.86 (0.34-2.20)
Women High normal: 130-139 /85-89 RR:1.19 (0.50-2.84)
Women Stage 1 HT: 140-159 /90-99 RR:2.02 (0.93-4.38)
Women Stage 2/3 HT: ≥160 /≥100 RR:4.09 (1.70-9.85)
2
3
Only RRs for men and women aged 40-59 given above. No HRs given
Sleight et al.,
2009546
CV events (CV
death, MI, HF,
Stroke)
SBP values (quartiles)
CV death
≤130 mmHg Reference
130-142 mmHg 0.98 (0.86-1.12)
142-154 mmHg 0.93 (0.81-1.06)
>154 mmHg 0.98 (0.86-1.11)
MI
≤130 mmHg Reference
130-142 mmHg 0.87 (0.74-1.01)
Hypertension (partial update)
Outcome
Okayama et al.,
2006466
Initiating and monitoring treatment, including blood pressure targets
Study
Study
Outcome
Stroke
≤130 mmHg Reference
130-142 mmHg 1.11 (0.92-1.33)
142-154 mmHg 1.32 (1.11-1.58)
>154 mmHg1.51 (1.28-1.79)
Mortality
(stroke, CHD
and all-cause)
SBP values
80-119 mmHg
120-129 mmHg
130-136 mmHg
137-149 mmHg
150-260 mmHg
No HRs given, nor any other RRs or ORs relevant to the categories above.
Fagard et al.,
2004208
CV events
Normal ABP: <140mmHg Reference
Abnormal ABP: 140-159mmHg RR: 1.27 (0.64-2.52)
High ABP: ≥160mmHg RR: 2.13 (1.09-4.13)
No HRs given, but unadjusted RRs above calculated from data in outcome table.
Gustavsen et al.,
2003 244
CV events
NT: <140; mean = 129.1 mmHg Reference
HT: SBP >140; mean = 160.3 mmHg HR p<0.001
WCH: CBP>140, mean = 136.3; ABPM <135/90 mmHg HR 6.6 (p<0.001)
HR p values given as shown, but no CIs and no HR value for HT were provided.
Inoue et al.,
Stroke
NT: <135 / <80 mmHg Reference
1
Weitzman et al.,
2006629
Hypertension (partial update)
CHF
≤130 mmHg Reference
130-142 mmHg 0.85 (0.71-1.01)
142-154 mmHg 0.87 (0.74-1.04)
>154 mmHg0.84 (0.71-0.99)
2
4
Initiating and monitoring treatment, including blood pressure targets
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
142-154 mmHg 0.88 (0.75-1.02)
>154 mmHg1.03 (0.88-1.20)
HF
Conen et al.,
2007136
Major
event
Deckers, 2006165
CV death
SBP values
NT (not on Tx) <120 mmHg Reference
120-129 mmHg 1.10 (0.89-1.37)
130-139 mmHg 1.35 (1.09-1.68)
HT (or on Tx) <130 mmHg 1.91 (1.36-2.68)
130-139 mmHg 2.61 (2.04-3.34)
140-149 mmHg 2.04 (1.63-2.55)
150-159 mmHg 2.66 (1.99-3.55)
≥160 mmHg 3.42 (2.33-5.04)
CV
Optimal: <120/ <75 0.51 (0.40-0.64)
Normal: 120-129/75-84 0.61 (0.48-0.76)
High normal: 130-139/85-89 Reference
HT: ≥140 /≥90 1.30 (1.08-1.57)
Age adjusted HR used
2
5
SBP values
≤130 mmHg
>130-160 mmHg
>160 mmHg
HRs not provided for above comparisons but multivariate HR for a 1mmHg increase in systolic BP: 1.01 (1.00-1.01)
Equiavlence studies
Table 30: Study details and results for equivalence studies determining thresholds for diagnosis and treatment using different blood
pressure measurement methods.
Reference
N
Population
Follow-up
Study design
BP values at baseline (groups / thresholds); mmHg
Clinic and ABPM measurements
Head et al.,
2010269
CLINIC MEASUREMENT CATEGORIES:
lower limits of grade 3 (severe) HT(180/110 mm Hg)
Hypertension (partial update)
Britton et al.,
2009101
Initiating and monitoring treatment, including blood pressure targets
Outcome
1
Study
2007285
HR (95% CI) for BP measurement (SBP/DBP)
[HRs given unless indicated. Available RRs or ORs have been given if no HRs available]
SDH: ≥135 / ≥80 mmHg 2.39 (1.48-3.87), p=0.0004
ISH: ≥135 / <80 mmHg 2.24 (1.33-3.76), p=0.0024
IDH: <135 / ≥80 mmHg excluded from model as number of subjects (n=37) and events (number not stated) were too low
Reference
N
8575
Population
NT and HT
Follow-up
Immediate
Author’s conclusions: equivalent thresholds
Clinic BP
threshold
ABPM predicted from staff
measured seated clinic BP
(n=5327)
ABPM predicted from doctor
measured seated clinic BP (n=1490)
24h
Night
Day
24h
Night
Day
Grade 3 (severe) HT
>180/110
163/101
157/93
168/105
151/95
143/86
155/98
Grade 2 (moderate) HT
>160/100
148/93
139/84
152/96
138/86
128/78
142/90
Grade 1 (mild) HT
>140/90
133/84
121/76
136/87
126/78
113/69
129/81
Target BP + 1 condition
<130/80
125/76
112/67
128/78
119/70
106/61
123/73
Target BP + proteinuria
<125/75
121/71
107/63
124/74
116/66
102/57
120/69
Normal BP
<120/80
117/76
102/67
120/78
113/70
99/61
117/70
Hypertension (partial update)
cross-sectional
study
BP values at baseline (groups / thresholds); mmHg
grade 2 (moderate) HT (160/100mmHg)
grade 1 (mild) HT (140/90 mm Hg);
for target upper limits for HT with associated conditions (130/80 mm Hg)
HT with substantial proteinuria (125/75 mm Hg
Upper limit of optimal normal (120/80 mm Hg).
1
2
6
Initiating and monitoring treatment, including blood pressure targets
Study design
ABPM equivalents
for clinic BPs
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
8.1.1
Evidence statements - clinical
Details of all the included studies are summarised in Table 31, Table 32 and Table 33.
 Most studies showed a continuous relationship between BP and risk of developing clinical
outcomes (ie. an increased risk of outcome with increasing BP value)
 This was true regardless of BP measurement method (office, ABPM, self-reported/ not
specified)
 The MA of Law et al.,351 showed that BP treatment reduced CVD risk regardless of pretreatment BP
 The Head 2010 study269 provided equivalent threshold values for ABPM and clinic BP
measurements for the diagnosis and treatment of HT.
Evidence statements – economic
No relevant cost-effectiveness evidence was identified.
Treatment of people aged 80 years and greater
Update 2011
127
Update 2011
Review question: in adults with primary hypertension, which is the most clinically and cost
effective first-line anti-hypertensive treatment (drug classes) in elderly people (aged ≥80
years)?
Clinical evidence
The literature was reviewed from December 2005 onwards (the cut-off date of the previous
guideline) for systematic reviews, RCTs and subgroup analyses of RCTs which addressed
first-line ant-hypertensive treatment in elderly people (aged ≥80 years) with primary
hypertension. Comparisons could be anti-hypertensive treatment or placebo. RCTs were
included if there was: ≥12 months follow-up and N≥200 (in accordance with the 2006
guideline criteria) and the population did not consist of people who were exclusively diabetic
or had CKD.
Two SR/MAs67,419 were found that fulfilled the inclusion criteria and addressed the question.
The first SR/MA (Musini et al 2009)419 was a Cochrane review and included N=8 studies.
The second SR/MA (Bejan-Angoulvant 2010)67 was an update of a previous SR/MA and
included additional data from the newer HYVET and HYVET-PILOT studies. , also consisted
of 8 studies in total, and was an update of the Cochrane SR/MA.
The Bejan-Angoulvant SR/MA67 was chosen to be included in this review instead of the
Cochrane SR/MA becauseit provided data for more outcome measures than the Cochrane
review, which pooled some outcomes together. Data was cross-checked between the two
SR/MAs.
The Began-Angoulvant SR/MA67 compared the development of clinical outcomes in patients
who were ≥80 years old who had been randomised to treatment with either anti-hypertensive
drugs or placebo. Data in the MA came from either sub-group analyses of RCTs (data from
only the ≥80 year-old people in the trial), or from RCTs in which only people ≥80 years were
enrolled. The mean follow-up time was 3.5 years (range 0 – 11.6) and the total number of
patients included was N=6701. The 8 included studies differed in terms of sample size, mean
SBP at baseline, follow-up time and the class of anti-hypertensive medication that patients
were randomised to in the active treatment arm (D, CCB or BB). However they were similar
in terms of the mean age of the study population (83 to 84 years old).
NOTE: The HYVET trial which was included in the MA, recruited people who were ‘less ill’
than those included in the other studies. Participants in HYVET were generally healthier than
those in the general population: they had low overall rates of stroke and death from any cause
and at basline they were generally free of multiple comorbid conditions (low prevalence of
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
previous cardiovascular disease, coronary artery disease and diabetes mellitus; inclusion
criteria also excluded people with heart failure, dementia or those requiring nursing care).
The evidence profile below (Table 31) summarises the quality of the evidence and outcome
data from the SR/MA included in this review,67 comparing treatment vs placebo in people
aged ≥80 years.
128
No of patients
No of
studies
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
anti-HT
treatment
Effect
Relative
Placebo
Quality
Absolute
(95% CI)
Mortality (all cause) (follow-up 0-11.6 years)
1
SR/MA
based on
8 RCTs*
no serious
limitations
no serious
inconsistency1,2
no serious
indirectness
serious3
none
1.06 (0.89,
1.25)
not enough
data given
in study to
calculate
MODERATE
data not given in study
0.83 (0.56,
1.22)
not enough
data given
in study to
calculate
LOW
data not given in study
0.65 (0.52,
0.83)
not enough
data given
in study to
calculate
HIGH
data not given in study
0.73 (0.62,
0.86)
not enough
data given
in study to
calculate
HIGH
Coronary events (follow-up 0-11.6 years)
12
9
1
SR/MA
based on
6 RCTs*
no serious
limitations
no serious
inconsistency
no serious
indirectness
very serious4
none
Stroke (follow-up 0-11.6 years)
1
SR/MA
based on
7 RCTs*
no serious
limitations
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
CV events (follow-up 0-11.6 years)
1
SR/MA
based on
6 RCTs*
no serious
limitations
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
Heart failure (follow-up 0-11.6 years)
Update 2011
data not given in study
Hypertension (partial update)
Summary of findings
Quality assessment
Initiating and monitoring treatment, including blood pressure targets
Table 31: Evidence profile comparing anti-hypertensive treatment versus placebo in people aged ≥80 years (systematic review/metaanalysis; Bejan-Angoulvant, 2010)67
NOTE: there was not enough data given in the study to calculate the HRs for these outcomes, so the RRs reported in the paper have been used
in the GRADE profile.
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
data not given in study
0.50 (0.33,
0.76)
not enough
data given
in study to
calculate
HIGH
data not given in study
0.99 (0.69,
1.41)
not enough
data given
in study to
calculate
LOW
data not given in study
0.80 (0.80,
1.11)
not enough
data given
in study to
calculate
MODERATE
data not given in study
0.98 (0.83,
1.15)
not enough
data given
in study to
calculate
VERY LOW
coronary death (follow-up 0-11.6 years)
1
SR/MA
based on
7 RCTs*
no serious
limitations
no serious
inconsistency
no serious
indirectness
very serious
4
none
Stroke death (follow-up 0-11.6 years)
1
SR/MA
based on
8 RCTs*
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious3
none
CV death (follow-up 0-11.6 years)
1
13
0
SR/MA
based on
8 RCTs*
no serious
limitations
serious
1
no serious
indirectness
very serious
4
none
*moderate quality SR/MA based on moderate and high quality RCTs
1
significant heterogeneity
2
NS heterogenity when HYVET trial removed
3
95% confidence interval includes both 1) no effect and 2) the MID (appreciable benefit or appreciable harm); or only just crosses the MID
4
95% confidence interval crosses both 1) no effect and 2) appreciable benefit or harm and non-appreciable benefit or harm
Hypertension (partial update)
no serious
limitations
Initiating and monitoring treatment, including blood pressure targets
1
SR/MA
based on
6 RCTs*
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
Economic evidence
One study (Szucs 2010580) was identified from the update search that examined the costeffectiveness of antihypertensive drug treatment in people over the age of 80 years. This is
summarised in the economic evidence profile below (Table 32, Table 33). A full evidence
table is also provided in Appendix G: Evidence tables – health economic studies (2011
update).
Table 32: Antihypertensive treatment versus no treatment in people aged over 80 years –
economic study characteristics
Study
Applicabilit
y
Szucs 2010580)
Switzerland
Partially
applicable(a)
Limitations
Other Comments
Potentially
serious
limitations(b)




HYVET study
Model based on HYVET RCT639
Time horizon: 2 years
Health outcomes: life years gained
Costs: antihypertensive drugs, acute management and
follow-up of MI, stroke and heart failure.
a) Some uncertainty about applicability of Swiss unit costs. QALYs not used. Discounting not in line with NICE reference
case.
b) Based on single RCT analysis and so does not incorporate all available evidence for patients over 80 years. Some
methodological issues about how health outcomes and costs are calculated and attributed in model.
Table 33: Antihypertensive treatment versus no treatment in people aged over 80 years –
economic summary of findings (mean per person)
580
Szucs 2010 )
Switzerland
Incremental
cost (£)
-£14(a)
HYVET study
Incremental
effects
0.0457 life
years gained
ICER
Uncertainty
Treatment dominated
no treated (lower
costs and improved
health outcomes)
One way sensitivity analyses of
20% variation in medication
cost, cost of stroke, cost of HF,
cost of MI, life expectancy.
Medication cost and cost of
stroke had the biggest impact.
Results varied from treatment
dominant to £1097 per life year
gained.
a) Converted from 2007 Swiss Francs.
Evidence statements – Clinical
Study data has come from one moderate quality systematic review/meta-analysis67 which
included eight moderate and high quality RCTs.
In people aged ≥80 years old, anti-hypertensive treatment was significantly better than
placebo for:
 stroke
[high quality evidence]
 CV events
[high quality evidence]
 heart failure
[high quality evidence]
There was NS difference between anti-hypertensive treatment and placebo in people aged ≥80
years old for:
 total mortality
[moderate quality evidence]
 coronary events
[low quality evidence]
 coronary death
[low quality evidence]
 stroke death
[moderate quality evidence]
131
Update 2011
Study
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
 CV death
[very low quality evidence]
Evidence statements – Health economic
 One partially applicable study with potentially serious limitations found treating people
over 80 years of age with hypertension was cost-effective compared to not treating them.
Link from evidence to recommendations
Two main sources of evidence informed the GDG discussion about blood pressure thresholds;
i) observational data examining the relationship between blood pressure and clinical outcomes
from normotensive and hypertensive people according to current threshold definitions, and ii)
studies examining the impact of treatment of hypertension on clinical outcomes, taking
account of the baseline and achieved blood pressure values in clinical trials. It was not
possible to pool data from these studies because they included people across varying age
ranges, at different levels of baseline cardiovascular risk and patients were either untreated or
treated with a range of medications that could have influenced cardiovascular disease risk and
clinical outcomes. Thus, studies were examined individually to determine the strength and
consistency of evidence to support recommendations for pharmacological treatment
thresholds and optimal blood pressure targets for people with treated hypertension.
A number of conclusions can be drawn from this analysis; i) there was a positive and
continuous relationship between baseline blood pressure levels and the subsequent risk of
clinical outcomes; ii) this relationship was consistent for the risk of stroke, ischaemic heart
disease, heart failure and cardiovascular mortality; iii) this increased risk was most strongly
related to systolic pressure, reflecting the fact that systolic pressure rises with ageing and most
studies are conducted in older rather than younger people; iv) there was a paucity of data and
no recent studies of the relationship between blood pressure and clinical events in younger
people, i.e. <40 years.
The GDG noted that clinical trials invariably recruited older patients at high cardiovascular
disease risk and that there were no trials that had been specifically designed to examine the
appropriate blood pressure thresholds for initiating pharmacological treatment
forhypertension. Nevertheless, the individual pharmacological treatment trials had usually
randomised people into studies based on systolic blood pressure thresholds of 140 or
160mmHg and diastolic pressure thresholds of 90 or 100mmHg. The GDG also discussed
whether recommending specific blood pressure treatment thresholds was justified. The GDG
noted that the results of a meta-analysis and systematic review of 248,445 people in 108
randomised controlled trials (Law et al) had shown that blood pressure lowering reduced the
risk of cardiovascular disease and stroke irrespective of the patients’ pre-treatment blood
pressure, even when pre-treatment pressures were as low as 110/70mmHg – suggesting that
blood pressure lowering treatment could be offered to any person at high risk of
cardiovascular disease, not just those with hypertension. The GDG concluded that such a
hypothesis was consistent with the continuous relationship between blood pressure and
clinical outcomes. However, it remainsl a hypothesis that requires prospective testing to
properly define the balance between efficacy and safety, especially in people with low
baseline blood pressure, as well as the cost-effectiveness of such a strategy.
With regard to treatment thresholds, the GDG agreed that the current grading of hypertension,
i.e. Stage 1 Hypertension (CBPM ≥140/90mmHg) or Stage 2 hypertension (CBPM≥160-100)
was useful to help stratify people for treatment and should be retained. Furthermore the GDG
could see no point in any further grading of hypertension beyond Stage 2 as it would have no
impact of treatment stratification or clinical decision making. In light of the fact that this
guideline update recommends using the ABPM daytime average BP to confirm the diagnosis
of hypertension for initiating treatment, it was necessary to define the ABPM daytime average
pressures that are equivalent to the thresholds for stages 1 and 2 hypertension, previously
defined according to CBPM readings alone. A large study of 8,575 (Head et al., 2010) 269
132
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
133
Update 2011
examined the equivalent Clinic blood pressure and ABPM day time average pressure for
normotensive and hypertensive people. Of interest, the difference between Clinic and ABPM
was greatest when measured by doctors in the clinic rather than other clinical staff. Based on
the clinic staff data, a mean daytime average ABPM of 136/76mmHg was equivalent to Stage
1 hypertension threshold defined according to a CBPM threshold of ≥140/90mmHg. The
136/76mmHg value was rounded to derive the threshold for defining stage 1 hypertension, i.e.
≥135/85mmHg according to the ABPM day time average. This ABPM diagnostic threshold is
similar to that used as the reference standard in the systematic review of the specificity and
sensitivity of the different blood pressure measurement methods for the diagnosis of
hypertension. The GDG concluded that an ABPM day time average of ≥135/85mmHg should
be used to define the threshold for Stage 1 hypertension.
In the study of Head et al,269 the current CBPM threshold for the diagnosis of Stage 2
hypertension, i.e. ≥160/100mmHg, was equivalent to an ABPM daytime average of
152/96mmHg, which the GDG rounded to 150/95mmHg. Thus, the GDG concluded that a
daytime ABPM average BP ≥150/95mmHg should be used to define the threshold for stage 2
hypertension.
In reviewing treatment thresholds, the GDG first reflected on the existing recommendation
(2004) that pharmacological treatment should be offered for stage 2 hypertension, i.e. when
the clinic blood pressure is ≥160-100mmHg (equivalent to an ABPM day time average of
≥150/95mmHg). This recommendation was based on the evidence review in 2004 which
suggested that this level of blood pressure alone was sufficient to convey sufficient risk to
benefit from pharmacological therapy for hypertension.The GDG reviewed this
recommendation alongside the current evidence review which reinforced the message of the
powerful effect of baseline blood pressure on clinical risk across a wide range of blood
pressures and that pharmacologic treatment of blood pressure at or above the stage 2
hypertension threshold was associated with a clinical benefits and a reduction in risk. The
GDG concluded that adults should be offered pharmacological treatment of hypertension at
stage 2 hypertension (ABPM daytime average blood pressure ≥150/95mmHg).
The GDG then discussed whether pharmacologic treatment should be offered to all adults
with Stage 1 hypertension, i.e. CBPM systolic pressure 140-159 and/or diastolic pressure 9099mmHg, and ABPM daytime averages of ≥135/85mmHg but <150/95mmHg. The existing
guidance from 2004 recognised the uncertainty about whether every adult with stage 1
hypertension should be offered treatment. The GDG noted that the current recommendation
is to offer treatment to some but not all people with stage 1 hypertension (2004). The
treatment being targeted at those with stage 1 hypertension and higher levels of
cardiovascular disease risk as indicated by the presence of one or more of; target organ
damage, established cardiovascular disease, the presence of concomitant disease that
increases cardiovascular disease risk such as diabetes or CKD, or in those whose 10 year
cardiovascular risk is estimated to be 20% or more (ref NICE CVD risk) 428.
The GDG discussed the fact that most of the people with stage 1 hypertension who would not
be offered treatment according to this guidance will be younger (i.e. <40 years) because of
their lower 10 year risk risk and lesser likelihood that they will have developed target organ
damage or have established cardiovascular disease. Furthermore, there maybe greater
uncertainty about the diagnosis of hypertension when blood pressure is close to the threshold
for stage 1 hypertension. The GDG concluded that pharmacological treatment should be
offered to people with stage 1 hypertension who also have higher levels of cardiovascular
disease risk as indicated by the presence of one or more of; target organ damage, established
cardiovascular disease, the presence of concomitant disease that increases cardiovascular
disease risk such as diabetes or CKD, or in those whose 10 year cardiovascular risk is
estimated to be 20% or more (ref NICE CVD risk)428. Moreover, those with stage 1
hypertension without any of these additional higher cardiovascular factors indicators, i.e.
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
uncomplicated stage 1 hypertension, would not usally be offered pharmacological therapy for
hypertension but; i) would be recomended to undertake lifestyle modifications (see section x),
and ii) should also be re-evaluated annually and pharmacological treatment offered if they
develop more severe hypertension, i.e. stage 2 hypertension, or they develop target organ
damage, diabetes, CKD, cardiovascular disease, or their estimated 10 year cardiovascular
disease risk rises to 20% or more. In reality, this means that most people with stage 1
hypertension will be offered pharmacologic treatment because age is a major determinant of
CVD risk and the majority of people with hypertension are older rather than younger.
However, the GDG discussed the dilemma created by this recommendation about what to
advise for younger people (i.e. <40 years) with “uncomplicated”stage 1 hypertension. This
dilema is created by the fact that younger people with stage 1 hypertension are less likely to
have overt evidence of target organ damage or vascular disease and assessment of their CVD
risk over a relatively short duration of 10 years is unlikely to adequately reflect their lifetime
risk of CVD. The GDG further discussed that this dilemma is compouned by the fact that
when compared with older populations; i) in younger people, the time course over which
clinical outcomes develop as a consequence of stage 1 hypertension are likely to be very long
and much longer then those encountered in conventional clinical outcome trials and
epidemiological studies. Thus, there is very much less epidemiological data linking
uncomplicated stage 1 hypertension in younger people with adverse clinical outcomes; ii)
younger people have not been included in clinical outcome trials in sufficient numbers to
evaluate the impact of the pharmacological treatment of stage 1 hypertension on clinical
outcomes and probably never will be as such trials would need to be unfeasibly large of too
long a duration to be practical; iii) 10 year CVD risk estimates are strongly age dependent and
as such, in younger people will rarely provide an indication for treatment of uncomplicated
stage 1 hypertension. The GDG concluded that uncomplicated stage 1 hypertension in
younger people is unlikely to be benign, blood pressure will most likely rise over time, and
that there is uncertainty surrounding whether delayed pharmacological treatment will
necessarily reverse any accumulated target organ or cardiovascular damage. The GDG also
discussed the need to develop more accurate estimates of the lifetime risk of younger people
with uncomplicated stage 1 hypertension and the cost-effectiveness of treatment. In this
regard, the GDG recognised the importance of thorough assessment of target organ damage to
exclude its presence before deciding not to offer pharmacological treatment of hypertension
for younger people with seemingly uncomplicated stage 1 hypertension – the GDG thus
recommended that evaluation of the potential benefit of treating uncomplicated stage 1
hypertension in younger people with regard to its impact on target organ structure and
function should be a priority for future research. Meantime, the GDG recommended that for
younger people (i.e. <40years) with uncomplicated stage 1 hypertension, specialist referral for
exclusion of secondary causes of hypertension (see section xx) and detailed evaluation of
target organ damage e.g. by echocardiography to exclude LVH and dysfunction, should be
considered before concluding not to offer treatment. Moreover, when treatment is not offered,
careful annual re-evaluation is necessary because blood pressure is likely to rise over time and
target organ damage may develop.
Recommendations
23. Offer antihypertensive drug treatment to people aged under 80 years with stage 1
hypertension who have one or more of the following:
 target organ damage
 established cardiovascular disease
 renal disease
 diabetes
134
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
 a 10-year cardiovascular risk equivalent to 20% or greater. [new 2011]
24. Offer antihypertensive drug treatment to people of any age with stage 2 hypertension.
[new 2011]
25. For people aged under 40 years with stage 1 hypertension and no evidence of target organ
damage, cardiovascular disease, renal disease or diabetes, consider seeking specialist
evaluation of secondary causes of hypertension and a more detailed assessment of potential
target organ damage. This is because 10-year cardiovascular risk assessments can
underestimate the lifetime risk of cardiovascular events in these people. [new 2011]
Recommendations for research
3. In people aged under 40 with hypertension, what are the appropriate thresholds for
intervention?
There is genuine uncertainty about how to assess the impact of blood pressure treatment in
younger people (aged under 40) with stage 1 hypertension, and no overt target organ damage
or CVD. In particular, whether those with untreated hypertension are more likely to develop
target organ damage and, if so, whether such damage is reversible. Target organ damage and
CVD as surrogate or intermediate disease markers are the only indicators that are likely to be
feasible in younger people because traditional clinical outcomes are unlikely to occur in
sufficient numbers over the time scale of a typical clinical trial. The data will be important to
inform treatment decisions for younger people with stage 1 hypertension who do not have
overt target organ damage.
8.2 Monitoring treatment efficacy
8.2.1
Review question: In adults with treated primary hypertension, what is the best method to
measure blood pressure (home vs ambulatory vs office) for response to treatment?
Clinical evidence
The literature was searched for all years and studies published since the original guideline
(2003 onwards) were included.
Two SRs/MAs96,290 and 3 RCTs137,439,554 were found that fulfilled the inclusion criteria and
assessed which was the best BP measurement method for monitoring treatment in order to
reach target BPs. All studies were of moderate to good quality. The first MA96 compared the
effects of home monitoring vs usual care on BP lowering and reaching BP targets. The second
MA290 compared BP measurements at end of treatment using office or home measurements.
The 4 RCTs all assessed the effects of home monitoring vs office or ABPM monitoring on BP
lowering and reaching BP targets.
NOTE: all RCTs were underpowered to detect a difference in BP. In order to detect a 5mm
difference, a sample size of N≥500 is needed.
The evidence profiles below ( Table 35,
Table 36, Table 37, Table 38 and Table 39) summarise the quality of the evidence and
outcome data from the studies included in this review.96,137,290,439,554.
135
Table 34: Evidence profile comparing self-monitoring vs. usual care (Bray 2010)96
Limitations
1 96
randomised
trials1
very
serious2
1 96
randomised
trials7
very
serious2
1 96
randomised
trials10
very
serious2
1 96
randomised
trials12
very
serious2
1 96
randomised
trials12
very
serious2
1
Meta-analysis of 20 RCTs
Unclear randomisation process; unclear allocation concealment; unclear blinding; unclear ITT analysis; unclear drop-out rates
3
I2 >50%
4
95% CI crosses MID
5
Not stated. Total number of patients was 5,898
6
p = 0.000
7
Meta-analysis of 23 RCTs
8
Not stated. Total number of patients was 6,038
9
p = 0.015
10
Meta-analysis of 12 RCTs
11
Not stated. Total number of patients was 2,260
12
Meta-analysis of 3 RCTs
13
Not stated. Total number of patients was 572
14
p = 0.89
15
p = 0.96
2
Table 35: Evidence profile comparing reduction in blood pressure using clinic and home measurements (Ishikawa 2008)290
Quality
VERY LOW
LOW
VERY LOW
LOW
LOW
Hypertension (partial update)
Design
Update 2011
Update 2011
136
No of
studies
Initiating and monitoring treatment, including blood pressure targets
Summary of findings
No of patients
Effect
Other
self
usual
Relative
Inconsistency
Indirectness
Imprecision
Absolute
considerations monitoring care (95% CI)
Change in clinic systolic blood pressure (mm Hg) (Better indicated by lower values)
no serious
3.82 lower (5.61 to 2.03
serious3
serious4
none
05
05
indirectness
lower)6
Change in clinic diastolic blood pressure (mm Hg) (Better indicated by lower values)
no serious
no serious
no serious
1.45 lower (1.95 to 0.94
none
08
08
inconsistency
indirectness
imprecision
lower)9
Proportion of patients achieving clinic blood pressure target
no serious
0/0 1.09 (1.02 to
3
serious
serious4
none
0/0 (0%)11
Not estimable
indirectness
(0%)11
1.16)6
Change in daytime ABPM systolic blood pressure (mm Hg) (Better indicated by lower values)
no serious
no serious
no serious
2.04 lower (4.35 lower to
none
013
013
inconsistency
indirectness
imprecision
0.27 higher)14
Change in daytime ABPM diastolic blood pressure (mm Hg) (Better indicated by lower values)
no serious
no serious
no serious
0.79 lower (2.35 lower to
none
013
013
inconsistency
indirectness
imprecision
0.77 higher)15
Quality assessment
Summary of findings
Quality assessment
Limitations Inconsistency
Indirectness
Imprecision
Other
considerations
Home blood pressure
measurement
randomised
trials1
very
serious2
3
05
-
MD 0 higher (0 to 0
higher)6
VERY
LOW
05
-
MD 0 higher (0 to 0
higher)7
VERY
LOW
Quality
Update 2011
Change in systolic blood pressure (mm Hg) (Better indicated by lower values)
no serious
1
serious
serious4
none
05
indirectness
Change in diastolic blood pressure (mm Hg) (Better indicated by lower values)
randomised
very
no serious
290
3
1
serious
serious4
none
05
trials1
serious2
indirectness
1
Meta-analysis of 22 RCTs. Data sets in which the methods of clinic BP measurements were not clearly described were excluded
2
Unclear randomisation process; unclear allocation concealment; unclear blinding; unclear ITT analysis; unclear drop-out rates
3
No details
4
Difference in change not stated
5
Not stated. Total number of patients was 6,322
6
Reductions in clinic and home SBP were: -14.7±0.04 and -11.8±0.04 respectively; p<0.001
7
Reductions in clinic and home DBP were: -10.7±0.03 and -8.1±0.05 respectively; p<0.001
290
Absolute
Clinic blood pressure
measurement
137
Table 36: Evidence profile comparing reduction in blood pressure using home and ambulatory measurements (Ishikawa 2008)290
Quality assessment
No of
studies
Design
1 290
randomised
trials1
very
serious2
no serious
inconsistency
1 290
randomised
trials1
very
serious2
no serious
inconsistency
1 290
randomised
trials1
very
serious2
no serious
inconsistency
Limitations Inconsistency
Summary of findings
No of patients
Relative
Other
Home blood pressure Ambulatory blood pressure
Indirectness Imprecision
(95%
considerations
measuerement
measurememnt
CI)
Change in daytime systolic blood pressure (mm Hg) (Better indicated by higher values)
no serious
no serious
none
03
03
indirectness
imprecision
Change in daytime diastolic blood pressure (mm Hg) (Better indicated by higher values)
no serious
no serious
none
03
03
indirectness
imprecision
Change in nighttime systolic blood pressure (mm Hg) (Better indicated by higher values)
no serious
no serious
none
03
03
indirectness
imprecision
Change in nighttime diastolic blood pressure (mm Hg) (Better indicated by higher values)
no serious
no serious
none
03
03
indirectness
imprecision
Absolute
Quality
MD 1.6 higher (1.1 to 2.2
higher)4
LOW
MD 0.2 higher (0.4 lower
to 0.8 higher)5
LOW
MD 3.8 higher (3.3 to 4.4
higher)4
LOW
MD 1.2 higher (0.6 to 1.8
higher)4
LOW
Update 20
randomised
very
no serious
trials1
serious2
inconsistency
1
Meta-analysis of 5 RCTs.
2
Unclear randomisation process; unclear allocation concealment; unclear blinding; unclear ITT analysis; unclear drop-out rates
1 290
Effect
Hypertension (partial update)
Design
Effect
Relative
(95%
CI)
Initiating and monitoring treatment, including blood pressure targets
No of
studies
No of patients
3
Not stated. Total number of patients was 801
p<0.001
5
p=0.55
4
Quality
VERY LOW
VERY LOW
LOW
138
LOW
VERY LOW
VERY LOW
VERY LOW
Update 2011
VERY LOW
Hypertension (partial update)
Summary of findings
No of patients
Effect
No of
Other
Home blood pressure
Ambulatory blood
Relative
Design Limitations Inconsistency Indirectness Imprecision
Absolute
studies
considerations
measurement
pressure measurement
(95% CI)
Home systolic blood pressure (mm Hg) (follow-up 24 weeks; Better indicated by lower values)
1 439 randomised
very
no serious
no serious
MD 2.6 higher (2.3 lower to
serious2
none
52
46
trials
serious1
inconsistency
indirectness
7.4 higher)3
Home diastolic blood pressure (mm Hg) (follow-up 24 weeks; Better indicated by lower values)
randomised
very
no serious
no serious
MD 2.6 higher (0.1 lower to
439
1
serious2
none
52
46
trials
serious1
inconsistency
indirectness
5.2 higher)4
24-h systolic blood pressure (mm Hg) (follow-up 24 weeks; Better indicated by lower values)
randomised
very
no serious
no serious
no serious
MD 0.6 higher (3.0 lower to
439
1
none
52
46
trials
serious1
inconsistency
indirectness
imprecision
4.3 higher)5
24-h diastolic blood pressure (mm Hg) (follow-up 24 weeks; Better indicated by lower values)
randomised
very
no serious
no serious
no serious
MD 1.5 higher (1.0 lower to
439
1
none
52
46
trials
serious1
inconsistency
indirectness
imprecision
3.9 higher)6
Clinic systolic blood pressure (mm Hg) (follow-up 24 weeks; Better indicated by lower values)
randomised
very
no serious
no serious
MD 1.1 higher (3.7 lower to
439
1
serious2
none
52
46
trials
serious1
inconsistency
indirectness
5.9 higher)7
Clinic diastolic blood pressure (mm Hg) (Better indicated by lower values)
randomised
very
no serious
no serious
MD 1.3 higher (5.0 lower to
439
1
serious2
none
52
46
trials
serious1
inconsistency
indirectness
2.3 higher)8
Number of patients who reached target BP (follow-up 24 weeks)
randomised
very
no serious
no serious
RR 1.33
143 more per 1000 (from 48
439
1
very serious9
none
30/52 (57.7%)
20/46 (43.5%)
trials
serious1
inconsistency
indirectness
(0.89 to 1.99)
fewer to 430 more)
Number of patients progressing to combination therapy (follow-up 24 weeks)
randomised
very
no serious
no serious
RR 0.97
20 fewer per 1000 (from 182
1 439
very serious9
none
34/52 (65.4%)
31/46 (67.4%)
trials
serious1
inconsistency
indirectness
(0.73 to 1.29)
fewer to 195 more)
1
Unclear allocation concealment; unclear blinding; no ITT analysis
2
95% CI crosses MID
3
p = 0.29
4
p = 0.06
5
p = 0.72
6
p = 0.23
7
p = 0.66
8
p = 0.46
9
95% CI crosses both MIDs
Quality assessment
Initiating and monitoring treatment, including blood pressure targets
Table 37: Evidence profile comparing treatment targeted to home DBP vs.treatment targeted to ambulatory DBP Niiranen 2006439
Table 38: Evidence profile comparing treatment managed with ambulatory measurements vs.treatment managed with clinic
measurements (Conen 2009)137
Effect
Quality
No of
studies
139
Other
Ambulatopry blood
Clinic blood pressure
Relative
Absolute
considerations pressure measurement
measurement
(95% CI)
Change in 24-h systolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
randomised
very
no serious
no serious
mean 3.6 lower (7.0 to 0.3
137
1
serious2
none
70
66
trials
serious1 inconsistency indirectness
lower)3
Change in 24-h diastolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
randomised
very
no serious
no serious
no serious
MD 0.9 lower (3.0 lower to
137
1
none
70
66
trials
serious1 inconsistency indirectness imprecision
1.1 higher)4
Change in clinic systolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
randomised
very
no serious
no serious
MD 4.4 lower (10 lower to
137
1
serious2
none
70
66
trials
serious1 inconsistency indirectness
1.1 higher)5
Change in clinic diastolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
randomised
very
no serious
no serious
no serious
MD 0.4 lower (3.6 lower to
137
1
none
70
66
trials
serious1 inconsistency indirectness imprecision
2.8 higher)6
Mean number of antihypertensive drugs used (follow-up 1 years; Better indicated by lower values)
randomised
very
no serious
no serious
mean 0.19 lower (0.53 lower
137
1
very serious7
none
70
66
trials
serious1 inconsistency indirectness
to 0.15 higher)8
Patients with controlled 24-h blood pressure (follow-up 1 years)
randomised
very
no serious
no serious
RR 1.41 (1.01 174 more per 1000 (from 4
137
1
serious2
none
42/70 (60%)
28/66 (42.4%)
trials
serious1 inconsistency indirectness
to 1.99)9
more to 420 more)
Patients with controlled office blood pressure (follow-up 1 years)
randomised
very
no serious
no serious
RR 1.19 (0.77 66 more per 1000 (from 80
1 137
very serious7
none
29/70 (41.4%)
23/66 (34.8%)
trials
serious1 inconsistency indirectness
to 1.83)10
fewer to 289 more)
1
No details on allocation concealment; open label; no ITT analysis
2
95% CI crosses MID
3
p = 0.03
4
p = 0.37
5
p = 0.12
6
p = 0.81
7
95% CI crosses both MIDs
8
p for difference = 0.49
p = 0.04
p = 0.4
10
Limitations Inconsistency Indirectness Imprecision
VERY LOW
LOW
VERY LOW
LOW
VERY LOW
VERY LOW
VERY LOW
Update 2011
9
Design
Hypertension (partial update)
No of patients
Initiating and monitoring treatment, including blood pressure targets
Quality assessment
Table 39: Evidence profile comparing treatment managed with home measurements vs.treatment managed with clinic measurements
(Staessen 2004)554
Effect
Quality
Design
Limitations
1 554
randomised
trials
serious1
no serious
inconsistency
1 554
randomised
trials
serious1
no serious
inconsistency
1 554
randomised
trials
serious1
no serious
inconsistency
1 554
randomised
trials
serious1
no serious
inconsistency
1 554
randomised
trials
serious1
no serious
inconsistency
1 554
randomised
trials
serious1
no serious
inconsistency
Update 2011
randomised
no serious
serious1
trials
inconsistency
1
Unclear allocation concealment
2
log-rank p<0.001
3
95% CI crosses MID
4
p <0.001
1 554
Other
Home blood pressure Clinic blood pressure
Relative
Absolute
considerations
measurement
measurement
(95% CI)
Patients able to permenantly stop antihypertensive drug treatment (follow-up 1 years)
no serious
no serious
RR 2.29 (1.45 144 more per 1000 (from 50
none
52/203 (25.6%)
22/197 (11.2%)
MODERATE
indirectness imprecision
to 3.63)2
more to 294 more)
Clinic systolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
MD 6.8 higher (3.6 to 9.9
serious3
none
203
197
LOW
indirectness
higher)4
Clinic diastolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
MD 3.5 higher (1.9 to 5.1
serious3
none
203
197
LOW
indirectness
higher)4
Home systolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
MD 4.9 higher (2.5 to 7.4
serious3
none
203
197
LOW
indirectness
higher)4
Home diastolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
no serious
MD 2.9 higher (1.5 to 4.3
none
203
197
MODERATE
indirectness imprecision
higher)4
24-h systolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
MD 4.9 higher (2.5 to 7.4
serious3
none
203
197
LOW
indirectness
higher)4
24-h diastolic blood pressure (mm Hg) (follow-up 1 years; Better indicated by lower values)
no serious
no serious
MD 2.9 higher (1.4 to 4.4
none
203
197
MODERATE
indirectness imprecision
higher)4
Inconsistency Indirectness Imprecision
Update 2011
140
No of
studies
Hypertension (partial update)
No of patients
Initiating and monitoring treatment, including blood pressure targets
Quality assessment
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
8.2.2
Economic evidence
An economic evaluation should ideally compare all relevant alternatives. No studies were
identified in the update search comparing all of clinic blood pressure monitoring (CBPM),
ambulatory blood pressure monitoring (ABPM) and home blood pressure monitoring
(HBPM) for assessing blood pressure (BP) control in treated patients.
Two studies comparing CBPM and ABPM in treated patients were identified but were
excluded as were judged to have serious methodological limitations374,512.
One study (Staessen 2004554) was identified that examined the examined the cost
effectiveness of HBPM compared with CBPM. This is summarised in the HBPM versus
CBPM economic evidence profile below (Table 40, Table 41). A full evidence table is also
provided in Appendix G: Evidence tables – health economic studies (2011 update). One other
study of this comparison was also identified but was excluded in line with the review protocol
as the HBPM included a telemonitoring component476. The Staessen 2004 study554 was also
included in the clinical review above. Note that this study is in a population diagnosed with
CBPM and this may impact the applicability to a population diagnosed by another method.
This is because if you are diagnosed by CBPM and then monitored by ABPM to some extent
the result will be about the people who were incorrectly diagnosed in the first place not just
differences in follow-up monitoring.
No cost-effectiveness studies were included in Clinical Guideline 18 relating to this topic.
Study
Staessen 2004
Belgium
554
Applicability
Limitations
Other Comments
Partially
applicable(a)
Potentially
serious(b)
 CBPM diagnosed population who are treated
or not treated.
 CPBM vs HBPM to assess BP control with
treatment intensified if DBP >89mmHg,
reduced if DBP <80mmHg.
 Within-RCT analysis.
 Costs: Antihypertensive drugs, physician
visits, HBPM.
a) Some uncertainty about applicability of Belgian resource use and unit costs. Some uncertainty about applicability to a
population not diagnosed with CBPM. QALYs not used (cost consequence analysis).
b) Given that blood pressure was significantly different, other clinical events and costs of these may be relevant and time
horizon may be insufficient. Within trial analysis and so does not incorporate all available evidence on differences
between options and results of this study inconsistent with meta analysis included in clinical review; clinical study
considered to have methodological limitations.No analysis of uncertainty.
Table 41: HBPM versus CBPM (assessing response to treatment) – economic summary
of findings (mean per person)
Study
Staessen 2004
Belgium
554
Incremental
cost (£)
Incremental
effects
-£256(a)
BP increased;
medication
discontinuation
increased; no
significant
difference in left
ventricular mass
or symptoms
ICER
Uncertainty
Lower costs
with HBPM
but worse BP
control
NR
a) Converted from 2002 Belgium 2002 using purchasing power parities468
Evidence statements – clinical
One well-conducted meta-analysis96 found that:
141
Update 2011
Table 40: HBPM versus CBPM (assessing response to treatment) – economic study
characteristics
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
 Self-monitoring was significantly better than usual care for:
o reducing clinic SBP and DBP (SBP: 20 RCTs, N=5898; DBP: 23 RCTs, N=6038) [very
low and low quality evidence]
o proportion of patients achieving target clinic blood pressure (12 RCTs, N=2260)
[very low quality evidence]
o There was NS difference between self-monitoring and usual care for reduction in mean
daytime SBP and DBP ABPM (3 RCTs, N=572).
[low quality
evidence]
 When self-monitoring was accompanied by an additional co-intervention, participants were
more likely to meet target blood pressure compared to when there was none.
One meta-analysis290 found that:
 with anti-hypertensive treatment (regardless of drug class used for treatment):
o clinic SBP and DBP fell significantly more than home blood pressure [very low quality
evidence]
– home blood pressure fell approximately 20% less than clinic blood pressure
– changes in clinic blood pressure were linearly related to those of home blood
pressure
– the difference between clinic blood pressure and homeblood pressure was
attributable to the difference in baseline blood pressure levels
o home blood pressure fell significantly more than daytime ambulatory SBP and nighttime ambulatory SBP and DBP
[low quality
evidence]
– daytime ambulatory SBP fell 15% less and night-time ambulatory SBP fell 30% less
than home blood pressure
o the reduction in daytime ambulatory DBP was NS different than the reduction in home
blood pressure
[low quality
evidence]
o changes in home SBP were intermediate between clinic and ambulatory SBPs (for 24h,
daytime and night-time measurements)
One RCT*439 found that there was NS difference between treatment targeted to home DBP vs.
targeted to ABPM DBP for:
 Home SBP and DBP blood pressure measurements (end of trial)
[very low quality
evidence]
 24h ABPM SBP and DBP blood pressure measurements (end of trial) [low quality
evidence]
 Clinic SBP and DBP blood pressure measurements (end of trial)
[very low quality
evidence]
 number of patients who reached target blood pressure
[very low quality
evidence]
 intensity of anti-hypertensive treatments (number of patients progressing to combination
therapy)
[very low quality
evidence]
One RCT137 found that:
 treatment managed with ABPM measurements was significantly better than treatment
managed with CBPM for:
142
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
143
Update 2011
8.2.3
o reductions in mean 24h ABPM SBP
[very low quality
evidence]
o number of patients with controlled 24-hour blood pressure [very low quality
evidence]
 there was NS difference between treatment managed with CBPM measurements versus
measured with ABPM for:
o reductions in mean clinic SBP and DBP
[low and very low quality
evidence]
o reductions in mean 24h ABPM DBP
[low quality evidence]
o number of patients with controlled clinic blood pressure measurements [very low
quality evidence]
o number of antihypertensive drugs used
[very low quality evidence]
554
One RCT* found that:
 treatment managed with home blood pressure was significantly better than treatment
managed with clinic blood pressure measurements for:
o number of patients who could permanently stop a-HT treatment
[moderate quality evidence]
 treatment managed with clinic blood pressure was significantly better than treatment
managed with home blood pressure measurements for :
o reduction in clinic SBP and DBP blood pressure
[low quality
evidence]
o reduction in home SBP and DBP blood pressure
[low and moderate
quality evidence]
o reduction in 24h ABPM SBP and DBP ABPM blood pressure
[low and moderate
quality evidence]
*NOTE: Both groups were given the same target BP for treatment, despite being measured by
the two different methods, which would lead to a systematic under-treatment in one of the
groups
Evidence statements – health economic
 No cost-effectiveness analyses were identified incorporating all of CBPM, ABPM and
HBPM in the assessment of response to treatment.
 One partially applicable study with potentially serious limitations found that in a
population diagnosed with hypertension using CBPM, monitoring response to treatment
and adjusting treatment using HBPM was cost saving compared to CBPM; blood pressure
control was however worse.
Link from evidence to recommendations
All clinical outcome trials have used CBPM to monitor treatment efficacy. Some of these
trials have embedded substudies using HBPM or ABPM to monitor treatment effects but for
the primary outcome measures, the blood pressure control was invariably monitored using
CBPM. A meta-analysis by Bray et al., 2010 96showed that patients self monitoring their own
blood pressure was associated with lower achieved CBPM and a greater liklihood of
achieving the clinic blood pressure target. Interestingly another analysis (Ishikawa aet al.,
2008)290 also found that HBPM averages fell approximately 20% less than the corresponding
CBPM but that the relationship between the two measures was linear. Two studies (Niiranen
et al., 2006 and Conen et al., 2009)137,439 examined whether monitoring blood pressure
control with CBPM versus ABPM or HBPM impacted on blood pressure control and the
number of treatements used to achieve the blood pressure targets and found no differences
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
26. Use clinic blood pressure measurements to monitor the response to antihypertensive
treatment with lifestyle modification or drugs. [new 2011]
27. For people identified as having a ‘white-coat effect’, consider ABPM or HBPM as an
adjunct to clinic blood pressure measurements to monitor the response to antihypertensive
treatment with lifestyle modification or drugs. [new 2011]
Research recommendations
4. In adults with primary hypertension, does the use of out-of-office monitoring (HBPM or
ABPM) improve response to treatment?
There is likely to be increasing use of home and ambulatory blood pressure monitoring for the
diagnosis of hypertension as a consequence of this guideline update. There are, however, very
little data regarding the utility of HBPM or ABPM as means of monitoring blood pressure
control or as indicators of clinical outcome in treated hypertension, compared with clinic
blood pressure monitoring. Studies should incorporate HBPM and/or ABPM to monitor blood
pressure responses to treatment and their usefulness as indicators of clinical outcomes.
8.3 Blood pressure targets for treatment
Review question: in adults with primary hypertension, what is the optimum BP that should be
reached for once treatment has been initiated/ targeted for treatment?
Clinical evidence
The literature was searched for studies published since the original guideline (2003 onwards).
All study types were included, if the population did not consist of people who were
exclusively diabetic or had CKD. Studies were excluded if they did not stratify results into
more than 1 different BP value / target.
Fifteen studies29,49,82,134,168,209,280,282,298,462,463,539,549,616,623,655 were found that fulfilled the
inclusion criteria and assessed what the optimum target blood pressure should be for treating
people with primary hypertension. One of the studies (29,298) was published as two separate
144
Update 2011
between blood pressure monitoring methods. The GDG noted that there was inadequate data
comparing the use of HBPM or ABPM to monitor blood pressure control and whether they
offer any important advantages over CBPM. Routine monitoring with HBPM or ABPM
would also require considerable investment in additional monitors beyond that required for
diagnosis of hypertension. The GDG recognised that patients may wish to monitor their own
blood pressure using HBPM and the possibility that engaging patients in their own blood
pressure monitoring process using HBPM could lead to better blood pressure control (NICE
Medicine’s Adherence Guideline, CG76)426. The GDG noted, however, that further data on
self-monitoring and self management of blood pressurewas required before this could be
recommended as the preferred modality for monitoring blood pressure control in people with
treated hypertension.
The GDG recommended that for people receiving antihypertensive medications, clinic blood
pressure readings should usually be used to monitor their response to treatment.
The GDG discussed how to monitor blood pressure in people with significant discrepancies
between their clinic blood pressure readings, recognising that CBPM may not provide an
accurate representation of their blood pressure control. In people identified as having a white
coat effect (people who are hypertensive according to their ABPM daytime average blood
pressure but with a CBPM at diagnosis that exceeded their ABPM by ≥20 mmHg systolic, or
≥10mmHg diastolic) the GDG recommended that HBPM should be considered as an adjunct
to CBPM to monitor the response to antihypertensive treatment and/or lifestyle modification.
Recommendations
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
papers reporting different assessment methods or outcomes, so this study has only been
counted once, however results from both papers are reported and referenced here.
The studies addressing the question were categorised into three different types:
1. More vs less intense treatment studies - (eight studies; eight
papers)29,82,280,282,298,463,549,616 – those that assess people who were randomised to more
intense (strict or intense) BP lowering vs. less intense (mild or standard) BP lowering
2. Within-treatment BP studies (eight studies)49,134,168,209,462,539,623,655 - those that assess
within-treatment / achieved BP values and the associated risk of developing clinical
outcomes.
3. Target BP studies(one study)462 - those that target people to different specific blood
pressure values (for example, according to age groups)
Details of all the included studies are summarised in Table 42 and Table 43 and Table 44.
NOTE: Data from the more vs less intense treatment studies was not pooled into metaanalysis because the studies varied widely in the following factors: treatment targets,
interventions used to reach the target (type of anti-hypertensive drug), follow-up times, BP
measurement method and outcome definitions. Therefore GRADE was performed on each
individual RCT to give a quality rating for each outcome measure used in the study (see
145
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
Table 45).
146
N
BPLTTC,
200882
190,60
6
SR/MA
31
RCTs
Populati
on
HT
not clear
if
underlyin
g diabetes
/ CKD
Clinic
HT
Home
173/104
(≥65
years)
14
7
Hosohata et al.,
2007280
971
165/104
(<65
years)
152/90
(more
and less)
Followup
HT
Clinic
172/89
(strict
and
Best Target BP
(authors’ conclusions)
QUALITY
Not
specified
(just more
vs. less
intense)
CV events
not reported
NS difference between
more vs. less intense BP
lowering regimens;
extent of risk reduction
was directly related to
the degree of BP
lowering
LOW and
VERY LOW
(age <65 and
>65
respectively);
based on
moderate
quality
SR/MA
which
included low
to high
quality RCTs)
12
months
More
intense
<125/80
BP
changes/ac
hievement
of target
BP
More: 132/80;
25%
Less: 133/79;
45%
NS difference between
more vs. less intense BP
lowering regimens for
change in BP; More
people in less intense
reached target BP.
MODERATE
AND LOW
BP
changes/ac
hievement
12 months:
Strict: 139/76;
Strict treatment group
was SS better for:
lower final BP value (1
MODERATE
Less
intense
125134/8084
4320
Outcomes
Final mean BP
(SBP/DBP
mmHg) and
number people
reaching target
Minimu
m of
1000
patient
years in
each trial
RCT (HOMEDBP)
JATOS study
group 2005 and
200829,298
Target
BP for
Treatme
nt (SBP /
DBP,
mmHg)
12
months
and 2
Strict
control
Update 2011
Reference /
study type
BP
measurem
ent
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
Hypertension (partial update)
Table 42: Study details and results for optimal blood pressure targets (trials comparing more vs. less intense blood pressure lowering
treatment regimens were used to assess this)
Initiating and monitoring treatment, including blood pressure targets
More versus less intense treatment studies
Populati
on
Followup
years
RCT (JATOS)
Mild
control
140-160
SBP
Outcomes
of target
BP;
morbidity
(CVD and
renal
failure)
and
mortality
Final mean BP
(SBP/DBP
mmHg) and
number people
reaching target
60%
Mild: 147/79;
67%
2 years:
Strict: 136/75
Mild: 146/78
Best Target BP
(authors’ conclusions)
and 2 years)
QUALITY
But was SS worse for
number of people
achieving target BP (1
year)
There was NS difference
for morbidity and
mortality at 2 years
14
8
Solomon et al.,
2010549
228
HT
Clinic
161/90
(intensiv
e)
24 weeks
Intensive
treatment
<130 SBP
RCT (EXCEED)
162/94
(standard
)
Standard
treatment
<140 SBP
BP
changes/ac
hievement
of target
BP
Intensive:
131/75
Standard:
137/80
Intensive: 46%
<130; 82% <140
More intense treatment
was SS better for:
lower final BP value
MODERATE
AND LOW
More intense treatment
increased chance of
achieving SBP <140
mmHg
Standard: 60%
<140
Verdecchia et
al., 2009616
RCT (CardioSis)
1111
HT
Clinic
163/90
(tight
and
usual
control)
2 years
Tight
control
<130 SBP
Usual
control
<140 SBP
BP
changes/ac
hievement
of target
BP;
CV
endpoint
Tight: 132/77
Usual: 136/79
Achieved <140:
Tight 79%
Usual 67%
Tight control group was
SS better for:
reduction in CV events
percentage achieving
SBP (<130 and <140)
reduction in BP value
MODERATE
Hypertension (partial update)
N
Target
BP for
Treatme
nt (SBP /
DBP,
mmHg)
<140 SBP
Initiating and monitoring treatment, including blood pressure targets
Reference /
study type
BP
measurem
ent
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
mild)
Populati
on
140
HT
RCT
Clinic
(pulse
pressure
analyser)
Followup
177/101
(mean)
12
months
Intense
control
<130/85
Final mean BP
(SBP/DBP
mmHg) and
number people
reaching target
Achieved <130:
Tight 72%
Usual 27%
Best Target BP
(authors’ conclusions)
BP
changes
Intense: 129/78
Moderate:
152/87
Intense control group
was SS better for:
reduction in BP value
LOW
BP
changes/ac
hievement
of target
BP;
CV
endpoint
Strict: 137/75
Moderate:
142/77
Strict control group was
SS better for:
percentage achieving
target BPs (<140 and
≥140 to <150)
reduction in BP value
MODERATE
AND LOW
Outcomes
QUALITY
Moderate
control
<140/90
14
9
Ogihara et al.,
2003463
3260
ISH
Clinic
169/81
(mean)
3.07
years
(median)
Strict
control
<140
RCT (VALISH)
Moderate
control
≥140 to
<150
mmHg
NT = normotensives; HT = hypertensives; ISH = isolated systolic hypertensives
78% and 48%
achieved target
(strict and
moderate groups
respectively)
There was NS difference
between the groups for::
reduction in CV events
Hypertension (partial update)
Ichihara et al.,
2003282
N
Target
BP for
Treatme
nt (SBP /
DBP,
mmHg)
Initiating and monitoring treatment, including blood pressure targets
Reference /
study type
BP
measurem
ent
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
Reference /
study type
Wang et al.,
2005623
SR/MA
15
0
Zanchetti et
al., 2009655
SR of
different
studies
N
12903
young (3049 years
≥160/95m
mHg) 3
trials;
14323 old
(60-79
years
≥160mmH
g/
<95mmHg)
5 trials;
1209 very
old patients
(≥80 years
≥160mmH
g/
<95mmHg)
a) low-risk
patients
(n=13
trials);
b) elderly
patients
(n=11
trials);
Population
HT
BP
measur
ement
method
Clinic
Baseline
mean
BP
(SBP/D
BP
mmHg)
young:
154/100
old:
174/83
very old:
176/78
HT (diabetic
studies
assessed by
subgroup
analysis)
Clinic
n/a
Followup
Median
young: 5
years;
old: 3.9
years;
very old:
3.8 years
Outcome
s
CV
events;
CV
mortality
In-treatment /
achieved BPs
young: ≥160 / ≥95
old and very old: ≥160
/ <95 (ISH)
Best Target BP (authors’
conclusions)
Anti-hypertensive treatment
improves outcomes mainly
by lowering SBP; Patients
with >median SBP reduction
risk of outcome decreased
regardless of decrease in
DBP or achieved DBP.
Active treatment tended to
reduce the risk of any
outcome to a similar extent
(i.e. DBP did not lead to
differences in cardiovascular
outcome as long as SBP
substantially decreased.
n/a
Total
mortality;
CV
events;
CV
mortality
Risk groups (High,
medium, low)
Achieved level of risk does
not appear to correlate
closely with the SBP values
achieved. In high risk
patients there is a ‘ceiling
effect’ for treatment
benefits. Delaying
therapeutic correction of CV
QUALITY
MODERATE
quality SR/MA
based on low
quality
observational
studies
MODERATE
quality SR/MA
based on low
quality
observational
studies
Hypertension (partial update)
Table 43: Study details and results for within-treatment / achieved blood pressure studies assessing the optimal blood pressure target
for treatment
Initiating and monitoring treatment, including blood pressure targets
Within-treatment blood pressure studies
RCT
(PROGRESS)
Treated as
observational
study as not
using
randomised
groups
Coca et al.,
2008134
Treated as
observational
study as not
using
randomised
groups
Population
6105
Cerebrovasc
ular disease
(not
necessarily
HT)
Clinic
22,576
HT
Clinic
Followup
Outcome
s
In-treatment /
achieved BPs
Best Target BP (authors’
conclusions)
risk factors until a high level
of risk is achieved,blunts the
full benefits of interventions.
Stratifie
d into:
<120;
120-139;
140-159;
≥160
Median
3.9 years
Risk of
Stroke
Stratified into:
<120; 120-139; 140159; ≥160
Patients with
cerebrovascular disease
would have lowest risk of
recurrence of stroke with BP
lowered to approximately
115/75mmHg
LOW
Stratifie
d into:
SBP
<140
vs. ≥140
61,836
patient
years
Fatal/nonfatal
stroke;
Achieving
target BP
<140/90
SBP Stratified into:
<140 vs. ≥140
Patients who achieved
follow up SBP <140mmHg
had lower risk of stroke than
those with SBP ≥140mmHg;
DBP <90mmHg had lower
risk than ≥90mmHg.
LOW
DBP:
DBP Stratified into:
<90 vs. ≥90
QUALITY
Hypertension (partial update)
15
1
Arima et al.,
200649
N
c) diabetic
patients
(n=11
trials; these
would be
outside our
inclusion
criteria);
d) high-risk
patients
(n=18
trials)
Initiating and monitoring treatment, including blood pressure targets
Reference /
study type
BP
measur
ement
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
Population
4583
HT
(systolic)
Clinic
26,512
HT
Clinic
Followup
Outcome
s
In-treatment /
achieved BPs
Best Target BP (authors’
conclusions)
Mean
174/86
median 2
years;
further 4
years+
followup
Cerebrova
scular
events;
CHD
events;
mortality;
CV
events;
CV
mortality
DBP Stratified into:
≥95; <9585; <85-75;
<75-65; <65-55; <55
Antihypertensive treatment
can be intensified to prevent
cardiovascular events when
systolic BP is not under
control in older patients with
systolic hypertension, at
least until diastolic BP
reaches 55mmHg, except in
patients with coronary heart
disease (MI/angina), in
whom diastolic should not
be lowered to <70mmHg.
LOW
Mean
166/95
Mean 3
years
Composit
e of CV
events
SBP Stratified into:
<130; 130-139; 140149; 150-159; ≥160
Clear relationship between
BP control and
cardiovascular events;
incidence of events
increased in patients with
SBP ≥140/85mmHg
(≥140/90mmHg in very
elderly) and in diabetic
patients with BP
≥130/85mmHg during
treatment. Results suggest
that BP should be below
140/90 for reducing the risk
of CV events. BP was
LOW
Post-hoc
analysis of
RCT (SystEur)
15
2
Treated as
observational
study as not
using
randomised
groups
Shimamoto et
al., 2008539
Within-group
comparison
study (JHEALTH)
DBP Stratified into:
<75; 75-79; 80-84;
85-90; ≥90
QUALITY
Hypertension (partial update)
RCT
(INVEST)
Fagard et al.,
2007209
N
Initiating and monitoring treatment, including blood pressure targets
Reference /
study type
BP
measur
ement
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
<90 vs.
≥90
Population
22,576
HT
Clinic
A-priori
subanalysis of
RCT
(INVEST)
15
3
Treated as
observational
study as not
using
randomised
groups
NT = normotensives; HT = hypertensives;
Overall
mean:
149.5/86
.3
Followup
Outcome
s
In-treatment /
achieved BPs
24
months
Mortality,
MI stroke
Stratified into agegroups and SBP / DBP
nadirs.*
Age
BP nadirs
SBP
DBP
<60
110
75
60<70
70<80
≥80
115
75
135
75
140
70
Best Target BP (authors’
conclusions)
controlled below 140.90
mmHg in the very elderly
patients (≥85 years) and they
also had a lower risk of CV
events.
J-shaped relationship
(among each age-group)
with on-treatment SBP and
DBP and clinical end-points
/ events. SBP at HR nadir
increased with increasing
age – highest for teh very
old (140 mmHg). DBP at
HR nadir was only slightly
loer for the very old (70
mmHg). Therefore optimal
management may involve a
higher target SBP and lower
target DBP for very old
people (≥80 years) vs other
age-groups.
QUALITY
LOW
Hypertension (partial update)
Denardo et al.,
2010168
N
Initiating and monitoring treatment, including blood pressure targets
Reference /
study type
BP
measur
ement
method
Baseline
mean
BP
(SBP/D
BP
mmHg)
Reference /
study type
Ogihara et al.,
2009462
15
4
Sub-analysis
of RCT
(randomised to
ARB vs ACEi)
treated as
observational
study as not
using
randomised
groups
N
4703
Populatio
n
HT
BP
measure
ment
method
Office
Baseline
mean
blood
pressure
(SBP/D
BP
mmHg)
Overall:
163/92
Followup
Mean 3.2
years
Outcomes
CV events
In-treatment /
achieved blood
pressure
All people: 136/78
Best Target blood pressure
(authors’ conclusions)
Higher achieved blood
pressure was associated with
increased risk of CV events.
QUALITY
LOW
Hypertension (partial update)
Table 44: Study details and results for target blood pressure studies assessing the optimal blood pressure target for treatment
Initiating and monitoring treatment, including blood pressure targets
Target BP studies
No of patients
No of
studies
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
more intense
BP lowering
less intense
BP lowering
Effect
Relative
Quality
Absolute
(95% CI)
CV events (aged <65 years): SR/MA - BPLTTC (follow-up 1000 patient-years)
1
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious2
none
212/5024
(4.2%)
365/9360
(3.9%)
RR 0.88
(0.75 to
1.04)
5 fewer per
1000 (from
10 fewer to 2
more)
LOW
260/4198
(6.2%)
RR 1.03
(0.85 to
1.24)
2 more per
1000 (from 9
fewer to 15
more)
VERY LOW
CV events (aged >65 years): SR/MA - BPLTTC (follow-up 1000 patient-years)
1
randomised
trials
serious
1
no serious
inconsistency
no serious
indirectness
very serious
3
none
156/2251
(6.9%)
Final home SBP 12 months (Hosohata 2007 study) (follow-up 12 months; measured with: mmHg; Better indicated by lower values)
15
5
1
randomised
trials
serious4
no serious
inconsistency
no serious
indirectness
serious5
none
817
870
-
MD 1 lower
(2.2 lower to
0.2 higher)6
LOW
392/870
(45.1%)
RR 0.44
(0.38 to
0.52)8
252 fewer
per 1000
(from 216
fewer to 279
fewer)
MODERATE
1453/2155
(67.4%)
RR 0.88
(0.84 to
0.92)8
81 fewer per
1000 (from
54 fewer to
108 fewer)
MODERATE
% reaching BP target (Hosohata 2007 study) (follow-up 12 months)
1
randomised
trials
serious
4
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
163/817
(20%)
% reaching BP target (JATOS study group) (follow-up 1 years)
1
randomised
trials
serious
9
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
1288/2165
(59.5%)
Change in SBP (JATOS study group) (follow-up 1 years; measured with: mmHg; Better indicated by lower values)
Hypertension (partial update)
Summary of findings
Quality assessment
Initiating and monitoring treatment, including blood pressure targets
Table 45: GRADE profile for more versus less intense treatment studies
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
2165
2155
-
MD 7.20
lower (8.05
to 6.35
lower)10
9/2165
(0.4%)
8/2155
(0.4%)
RR 1.12
(0.43 to
2.9)11
0 more per
1000 (from 2
fewer to 7
more)
MODERATE
86/2155 (4%)
RR 1.0
(0.74 to
1.33)11
0 fewer per
1000 (from
10 fewer to
13 more)11
MODERATE
none
MODERATE
Mortality (JATOS study group) . (follow-up 2 years)
1
randomised
trials
serious9
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
Morbidity (JATOS study group) (follow-up 2 years)
1
randomised
trials
serious
9
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
86/2165 (4%)
Change in SBP (Solomon 2010) (follow-up 2 years; measured with: mmHg12; Better indicated by lower values)
1
randomised
trials
serious13
no serious
inconsistency
no serious
indirectness
serious5
15
6
114
114
-
MD 5.30
lower (0 to 0
higher)
LOW
94/114
(82.5%)
68/114
(59.6%)
RR 1.38
(1.16 to
1.64)14
227 more
per 1000
(from 95
more to 382
more)
MODERATE
399/507
(78.7%)
334/499
(66.9%)
RR 1.18
(1.09 to
1.27)10
120 more
per 1000
(from 60
more to 181
more)
MODERATE
27/507
(5.3%)
52/499
(10.4%)
HR 0.50
(0.31 to
0.79)16
51 fewer per
1000 (from
21 fewer to
71 fewer)
MODERATE
none
% reaching target (Solomon 2010) (follow-up 2 years)
1
randomised
trials
serious13
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
% reaching target (Verdecchia 2009) (follow-up 2 years)
1
randomised
trials
serious
15
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
CV events (Verdecchia 2009) (follow-up 2 years)
1
randomised
trials
serious
15
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
Hypertension (partial update)
serious
9
Initiating and monitoring treatment, including blood pressure targets
1
randomised
trials
randomised
trials
serious15
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
399/507
(78.7%)
none
334/499
(66.9%)
120 more
per 1000
(from 60
more to 181
more)
MODERATE
-
MD 23
lower (0 to 0
higher)19
LOW
MD 5.40
lower (6.31
to 4.49
lower)10
LOW
Final SBP (Ichihara 2003) (follow-up 2 years; measured with: mmHg; Better indicated by lower values)
1
randomised
trials
very serious18
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
71
71
Change in SBP (Ogihara 2010) (follow-up 2 years; measured with: mmHg; Better indicated by lower values)
1
randomised
trials
serious15
no serious
inconsistency
no serious
indirectness
serious5
none
1545
1534
-
0/1545 (0%)
0/1534 (0%)
RR 1.41
(1.33 to
1.5)10
47/1545 (3%)
52/1534
(3.4%)
HR 0.89
(0.6 to
1.31)11
% reaching target (Ogihara 2010) (follow-up 2 years)
1
randomised
trials
serious15
no serious
inconsistency
no serious
indirectness
no serious
imprecision7
none
MODERATE
15
7
CV events (Ogihara 2010) (follow-up 2 years)
1
1
randomised
trials
serious15
no serious
inconsistency
no serious
indirectness
RCTs included were of low to high quality; the SR/MA itself was of moderate quality
95% CI crosses both no effect and the lower MID (appreciable benefit/harm)
95% CI crosses both MIDs (appreciable benefit and appreciable harm)
4
randomised, ITT, but underpowered and attrition bias
5
95% CI crosses the lower MID
6
NS difference between groups
7
95% CI does not cross either MID
8
Favours less intense (p<0.00001)
9
Unclear allocation concealment
10
Favours Intense (p<0.00001)
11
p>0.05 (NS)
12
Favours intense (p=0.03)
13
open label, not true ITT
14
Favours intense (p=0.0002)
15
Inadequate allocation concealment and blinding
16
Favours intense (p=0.03)
17
Favours intense (p<0.001)
18
single blind, inadequate allocation concealment, ITT unclear
19
Favours intense (p<0.05)
2
3
no serious
imprecision7
none
4 fewer per
1000 (from
13 fewer to
10 more)
MODERATE
Hypertension (partial update)
1
RR 1.18
(1.09 to
1.27)17
Initiating and monitoring treatment, including blood pressure targets
Change in SBP (Verdecchia 2009) (follow-up 2 years)
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
Health economic evidence
One study (Jonsson 2003308) was identified from the update search that compared different
blood pressure targets. This is summarised in the economic evidence profile below (Table 46,
Table 47). A full evidence table is also provided in Appendix G: Evidence tables – health
economic studies (2011 update). No cost-effectiveness studies were included in Clinical
Guideline 18 relating to this topic.
Table 46: Treatment targets – economic study characteristics
Comparators
Jonsson 2003
Sweden
Target DBP
<90mmHg
Target DBP
<85mmHg
Target DBP
<80mmHg
HOT study
Applicabilit
y
Partially
applicable(a)
Limitations
Other Comments
Potentially
serious(b)
 Within RCT analysis (HOT260).
 Population: Hypertension and DBP110115mmHg
 Follow-up: mean 3.8year.
 Costs: antihypertensive drugs, healthcare
visits, side effects, cardiovascular
hospitalisations.
a) Some uncertainty about applicability of international resource use and Swedish unit costs. QALYs not used (clinical
outcomes reported as not significantly different). Discounting not applied.
b) Within RCT analysis and so does not incorporate all available evidence on differences between targets; issues raised
with interpretation of clinical trial as achieved BPs very similar despite different targets.
Update 2011
Study
Table 47: Treatment targets – economic summary of findings (mean per person)
Incremental
cost (£)
Incremental
effects
ICE
R
Target DBP
<90mmHg
Reference
N/a
Target DBP
<85mmHg
£82(a)
Target DBP
<80mmHg
£181 (a)
Clinical outcomes
were reported as
not significantly
different between
groups – see
clinical evidence
review for
details260.
Study
Comparators
Jonsson 2003
Sweden
HOT study
Uncertainty
Differences in cost were
statistically significant
(p<0.01).
A sensitivity analysis
including non-CV
hospitalisations increased
total costs but differences
between groups were
similar.
a) Converted from 1995 Swedish Kroner.
158
Update 2011
Evidence statements – clinical
More vs. less intense treatment studies (moderate and low quality evidence) showed:
 NS difference for:
o CV events (2 studies)82,463 – RRR was related to degree of blood pressure lowering
o Change in blood pressure (1 study)280
o Morbidity and mortality (1 study)29,298
 Less intense was better for:
o More people reaching target (2 studies)29,280,298
 More intense was better for:
o Lower final blood pressure value (5 studies)29,282,298,463,549,616
o Reduction in CV events (1 study)616
o Percentage reaching target SBP <130 (1 study)616
o Percentage reaching target SBP <140 (3 studies)463,549,616)
In-treatment / achieved blood pressure studies showed that:
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
8.3.1
 Higher achieved blood pressure was associated with increased risk CV events (2 studies
and 1 SR/MA)168,539,623
 Achieved SBP did not correlate with risk CV events (1 SR/MA)655
 Blood pressure <140/90 had a lower risk of CV events (2 studies)134,539
 Lowest risk of stroke was at blood pressure 115/75 mmHg (1 study)49
 DBP did not lead to risk differences as long as SBP substantially decreased (1 SR/MA)655
 DBP <90 had a lower risk of stroke (1 study)134
 Up to DBP 55 (had lower risk of stroke) when SBP was controlled; except for MI/angina
patients where DBP should not be <70 (1 study)209
 Optimal management may involve a higher target SBP and lower target DBP for very old
people (≥80 years) vs other age-groups (1 study)168)
Target blood pressure studies showed that:
 Higher achieved blood pressure was associated with increased risk CV events (1 study)462
Evidence statements – economic
 One partially applicable within RCT analysis (HOT) with potentially serious limitations
found that lower blood pressure targets were associated with higher costs and no
significant difference in clinical outcomes.
Link from evidence to recommendations: blood pressure treatment targets.
The GDG assessed a series of studies to define optimal treatment targets for people receiving
antihypertensive therapy. The studies addressing this question were categorised into three
different types; i) meta-analyses/systematic reviews of trials that had examined “more versus
less” blood pressure lowering on treatment, i.e. people randomised to more intense versus less
intense blood pressure lowering; ii) analyses of the relationship between achieved blood
pressure on treatment versus clinical outcomes; iii) studies targeting patients to specific blood
pressure values.
The more versus less studies studies provided more robust evidence for treatment targets
because they are randomised controlled trials whereas the studies using post-hoc stratifaction
of on-treatment achieved blood pressures versus outcomes are not randomised and are
potentially confounded by the fact that the blood pressure response to treatment may reflect
underlying vascular damage, i.e. those responding less well to treatment may have more
underlying vascular damage and by inference a higher risk of clinical outcomes. Moreover,
such studies did not usually adjust the results according to baseline blood pressure, age and
other key variables. The results of the more versus less treatment studies failed to show a
consistent benefit of the lower blood pressure target on clinical outcomes82,463 but the relative
risk reduction did appear to be related to the extent of blood pressure lowering across the
range. One study 29,298 did show a benefit of more intensive lowering on cardiovascular
morbidity and mortality. More intensive blood pressure lowering, not surprisingly, was
associated with more patients reaching a lower final blood pressue value. One smaller study
(Verdechia etal., 2009)616 did show better regression of LVH with more intensive BP
lowering and also as a secondary analysis, a reduction in a composite of cardiovascular
outcomes. In studies randomising patients to less intensive blood pressure lowering, more
patients achieved the less intensive blood pressure target29,280,298 reflecting the fact that lower
blood pressure targets are more diifuclt to achieve and generally required more medications.
In two studies (one a systematic review) examining the impact of achieved blood pressure on
treatment versus clinical oucomes, a higher achieved blood pressure was associated with a
higher risk of cardiovascular events 168,539,623 and a blood pressure on treatment of
<140/90mmHg associated with a lower risk of cardiovascular events in two studies134,539.
Similarly, in one study, a higher achieved blood pressure was associated with a increased risk
159
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
160
Update 2011
cardiovascular events 462. In constrast, in one systematic review, the achieved systolic blood
pressure did not correlate with the risk of cardiovascular events (1 SR/MA)655. The risk of
stroke appeared particularly sensitive to achieved blood pressure on treatment with the lowest
risk in those with the lowest on-treatment blood pressure, down to a value of 115/75 mmHg
49
. Similar findings were observed for on-treatment stroke risk in the analysis of Sleight et al
(2009). This latter study also stratified on treatment outcomes according to baseline blood
pressure and showed that those in patients with a baseline systolic blood pressure
<130mmHg, further blood pressure lowering appeared to be associated with an increased risk
of cardiovascular events. This latter finding from a large clinical trial of patients at high
cardiovascular risk does not support the uncritical adoption of lowering blood pressure in all
patients at high risk of cardiovascular disease, irrespective of their baseline blood pressure.
A Cochrane analysis of prospective studies of more versus less blood pressure treatment
identified only studies randomised on the basis of lowering diastolic pressure and showed no
evidence of more versus less blood pressure lowering on clinical outcomes (add ref – we did
discuss). The same analysis noted an absence of any studies designed to prospectively
examine the optinal systolic treatment target.
A formal cost effectiveness analysis of more versus less blood pressure lowering was not
prioritised as there was no clear evidence of effectivenss. From this perspective, one
potentially applicable study was identified (HOT study)260 with potentially serious limitations.
This study found that lower blood pressure targets were associated with higher costs, due to
the requirement for more treatment and no significant difference in clinical outcomes.
Based on these analyses, the GDG concluded that most clinical trials had adopted a treatment
target of <140/90 mmHg and that there was no convincing evidence supporting a lower
treatment target for the pharmacological treatment of hypertension. That said, the evidence
specifically examining optimal treatment targets for hypertension is inadequate and
consequently the optimal treatment target could not be clearly defined with certainty. The
GDG recommended that the target blood pressure for people treated for hypertension should
be <140/90 mmHg (consistent with the usual target bloodpressure in clinical outcome trials),
based on clinic blood pressure readings. For those with a white coat effect and thus requiring
HBPM to monitor their blood pressure control, or those patients preferring to use HBPM to
monitor their blood pressure control, the recommended target should be a HBPM average of
<135/85mmHg (based on the equivalent values for CBPM versus HBPM used for diagnosis
of hypertension). The GDG also noted the need for further studies prospectively randomising
people to more versus less systolic blood pressue lowering to determine the optimal systolic
pressure treatment target for people with treated hypertension.
Blood pressure thresholds and targets for people over the age of 80 years:
Previous guidelines in 2004 and 2006 noted the considerable uncertainty surrounding the
balance of benefits and risk when considering initiating blood pressure lowering treatment for
people over the age of 80 years. The uncertainty reflected tha fact that people over the age of
80 years had largely been excluded from recruitment into blood pressure treatment trials and
thus, the evidence of benefit of treatment in this age group had not been established. Whilst it
seemed likely that these people would accrue benefits from blood pressure lowering, it was
also conceivable that treatment coud lead to more adverse effects such as syncope and falls,
that might have offset any benefits of treatment.
The GDG considered one systematic review (Bejan-Angoulvant, 2010)67 which compared the
development of clinical outcomes in people aged ≥80 years who had been randomised to
antihypertensive treatment versus placebo. This meta-analysis included data from 8 studies,
including subgroups aged ≥80 years who had been randomized into treatment trials as well as
one large study (HYVET study) (Beckett, et al 2009)63 which included only hypertensive
people aged ≥80years. The total sample size was 6,701 and the mean follow-up was 3.5 years.
The baseline blood pressure and initial therapy differed between studies. The results of the
Hypertension (partial update)
Initiating and monitoring treatment, including blood pressure targets
analysis showed that in hypertensive people ≥80 years, pharmacological treatment was
significantly better than placebo for reducing the risk of stroke, cardiovascular events and
heart failure. The HYVET study provided the most robust and highest quality evidence and
had randomised people at a clinic systolic blood pressure threshold of ≥160mmHg and
treated blood pressure to a clinic blood pressure target of <150/90mmHg. The GDG noted
that the population randomised into the HYVET study were generally healthier, with lower
comorbidity than typically seen in this age group.
The GDG recommended that people aged ≥80 years, should be offered pharmacological
treatment for hypertension when they have stage 2 hypertension, i.e. when their ABPM
daytime average blood pressure is ≥150/95mmHg and should be treated to a clinic blood
pressure target of <150/90mmHg. If HBPM is being used to monitor blood pressure control in
people over the age of 80 years, then the blood pressure target equivakent to the
recommended CBPM target of <150/90mmHg, using a HBPM average would be
~140/85mmHg.
This recommendation regarding the treatment of people over the age of 80 years applies to
people who have stage 2 hypertension but are not currently treated when they reach the age of
80 years. It does not mean that people reaching this age who have been previously treated at
lower levels of blood pressure and/or to a lower treatment target of <140/90mmHg should
have their treatment back-titrated. There is an important distinction between continuing longterm and well-tolerated treatment in people over the age of 80 years and the initiation of blood
pressure lowering therapy at that age. For the latter, the evidence supports initiation of
treatment at stage 2 hypertension, treating to a CBPM target of <150/90mmHg. It is
conceivable lower thresholds and targets for this age group might be appropriate, however,
the balance if safety and efficacy for a more aggressive treatment strategy has not been
established. Indeed, before the emergence of the recent evidence (see above), there was
genuine uncertainty about the balance of efficacy versus harm with regard to initiating blood
pressure treatment in people aged 80 years or over. In this regard, the GDG also noted that the
key studies supporting this recommendation generally included older people who were fit and
active and had low levels of comorbidities. The GDG recommended that treatment decisions
in those aged ≥80 years should be based on the realistic expectations of clinical benefit from
treatment in the context of other comorbidities which might limit life expectancy.
Furthermore, the GDG recommended that for older patients who are already receiving
antihypertensive treatment when they reach the age of 80 years, the aforementioned evidence
supports continuation of treatment.
Recommendations
29. Aim for a target clinic blood pressure below 150/90 mmHg in people aged 80 years and
over with treated hypertension. [new 2011]
30. When using ABPM or HBPM to monitor the response to treatment (for example, in
people identified as having a ‘white-coat effect’ and people who choose to monitor their
blood pressure at home) aim for a target average blood pressure during the person’s usual
waking hours of:
 below 135/85 for people aged under 80 years
 below 145/85 in people aged over 80 years and over. [new 2011]
161
Update 2011
28. Aim for a target clinic blood pressure below 140/90 mmHg in people aged under 80 years
with treated hypertension. [new 2011]
Hypertension (partial update)
Integrating the assessment of blood pressure, target organ damage and
cardiovascular risk assessment and clinical decision making regarding treatment
initiation, treatment and targets
Research Recommendation
8.4 Frequency of review
Antihypertensive medications are used extensively to manage hypertension; dose titrations,
symptoms and blood pressure need to be managed and monitored. The guideline development
group affirms the importance of fully involving patients in prescribing decisions and
supporting them when starting, increasing, reducing or ceasing medicine to promote safety, a
good health outcome and patient satisfaction. Periodic review of medicines, lifestyle and
patient values and circumstances is thus an important aspect of good patient care. Although
there is no evidence for the optimal period, the guideline development group felt that face-toface medication review should occur once a year as a minimum to provide advice, review
symptoms and revise medication when appropriate.
Update 2011
5. In people with treated hypertension, what is the optimal systolic blood pressure?
Data on optimal blood pressure treatment targets, particularly for systolic blood pressure, are
inadequate. Current guidance is largely based on the blood pressure targets adopted in clinical
trials but there have been no large trials that have randomised people with hypertension to
different systolic blood pressure targets and that have had sufficient power to examine clinical
outcomes.
Integrating the assessment of blood pressure, target organ damage
and cardiovascular risk assessment and clinical decision making
regarding treatment initiation, treatment and targets
162
Update 2011
The algorithms found in Section 5.1 illustrate the recommended schema for the assessment of
blood pressure, clinical decision making regarding initation of treatment and review. Clinic
blood pressure is usually measured at scheduled reviews in primary care or on occasions
opportunistically during health screening. When clinic blood pressure is <140/90mmHg,
further investigation is not usually indicated and clinic blood pressure should be re-measured
at least every five years. More frequent review should be considered in people whose clinic
blood pressure is close to the 140/90mmHg threshold or in those in whom there is evidence of
cardiovascular disease or when their estimated 10 year cardiovascular disease risk is close to,
or exceeds 20%.
People with a clinic blood pressure ≥140/90mmHg should be offered ABPM to determine
whether their daytime ABPM average is ≥135/95mmHg. If a person’s ABPM daytime
average is <135/85mmHg they should be offered annual review. If the ABPM daytime
average is ≥135/85mmHg (i.e. stage 1 hypertension), they should be offered lifestyle advice
and considered for pharmacological treatment. If their ABPM day time average is
≥150/95mmHg (i.e. stage 2 hypertension), they should be offered lifestyle advice and
pharmacological treatment.
All people considered hypertensive should undergo routine clinical evaluation to determine
the presence of target organ damage, cardiovascular disease, diabetes or CKD and have their
10 year cardiovascular disease risk estimated. A review of lifestyle factors that may contribute
to the development of hypertension and/or increase a patient’s cardiovascular disease risk
should also be undertaken. If the initial clinical evaluation suggests the possibility of
secondary hypertension, the patient should be referred for specialist review.
If the patient has stage 1 hypertension and evidence of TOD, cardiovascular disease, diabetes,
CKD, or their estimated 10 year CVD risk is ≥20%, they should be offered treatment. If not,
they should be offered lifestyle advice and annual review as their blood pressure and
Hypertension (partial update)
Lifestyle interventions
9
Lifestyle interventions
9.1 Overview
A vast epidemiological literature describes an apparent relationship between raised blood
pressure and lifestyle choices and habits. For example, observational studies have shown that
people with raised blood pressure tend also to have low dietary calcium627. Does inadequate
intake of dietary calcium promote raised blood pressure or is the relationship a spurious one,
arising from inadequate adjustment for other hard-to-measure influences (a common problem
in observational studies). There is similar controversy about the role of diet, exercise, alcohol,
caffeine, potassium and magnesium supplements, sodium (table) salt and relaxation therapies.
Cause and effect can only be established by repeated and methodologically sound randomized
controlled trials, supported by evidence of a plausible biological mechanism, particularly
when the potential benefit is small.
Randomized controlled trials, enrolling patients who had raised average blood pressure
defined as systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥85 mmHg,
analysing either blood pressure or major cardiovascular endpoints on an intention-to-treat
basis, of eight weeks or more follow-up, are included in this review. However, none of the
studies identified were designed to quantify significant changes in rates of death or
cardiovascular events due to lifestyle interventions: instead they relied on the surrogate
endpoint of reduced blood pressure with its epidemiological link to reduced rates of disease.
Thus the evidence is less direct than for drug interventions which show reductions in
morbidity directly. The requirement that trials have a follow-up of at least eight weeks is
arbitrary but it reflects the belief that shorter time frames cannot usefully inform us about
enduring changes in blood pressure.
We searched electronic databases (Medline, Embase, CENTRAL) from 1998 to July 2003 for
reports of relevant randomised controlled trials; articles published before 1998 were identified
from hypertension guidelines, systematic reviews and meta-analyses31,118,187,192,214,293,366,388,
37,117,153,204,205,238,239,248,251,268,279,299,300,319-323,444,489,632-634 152,241,350,407
,
. Though there were a
number of trials informing most of the areas of interest, the trials were commonly small and
163
Update 2011
cardiovascular disease risk will increase over time. For younger people i.e. aged <40 years,
special consideration should be given to the possibility of secondary hypertension and the
exclusion of target organ damage before deciding not to initatite therapy for stage 1
hypertension and specialist review should be considered. If not offered pharmacological
treatment, they should be offered lifestyle advice and annual review.
If the initial clinic blood pressure is ≥180/110mmHg and there is evidence of target organ
damage and/or cardiovascular disease, the initiation of pharmacological therapy should not be
delayed whilst awaiting the results of ABPM. If the initial evaluation suggests the possibility
of accelerated hypertension or phaechromocytoma, the patient should be referred immediately
(same day) for specialist care.
When pharmacological treatment is considered, all patients should be offered lifestyle advice
(see section 9). People at higher risk, i.e. with target organ damage, established CV disease,
diabetes, CKD or an estimated 10 year CVD risk ≥20%, should be considered for additional
therapy to reduce their cardiovascular disease risk (e.g. statins and antiplatelet therapy) if not
already initiated (see NICE guidance on CVD risk, statins and antiplatelet therapy).
When pharmacological treatment is offered, clinic blood pressure should usually be used to
monitor the response to treatment and the target blood pressure is <140/90mmHg in people
aged <80 years and <150/90mmHg in people aged ≥80 years.
For people with white coat hypertension (see section 5.6), home blood pressure monitoring
(section 8.2) should be considered to monitor the response to treatment - the target blood
pressure for optimal treatment is a HPBM average of <135/85mmHg.
Hypertension (partial update)
Lifestyle interventions
the intervention of short duration (several months) relative to the progression of raised blood
pressure and cardiovascular disease. The quality of reporting of studies was commonly poor
(Table 48) and this may reflect poor methodological conduct, further weakening the strength
of evidence and consequent recommendations for clinical care.
164
Number of
studies
Number of
participants
Diet
14
Exercise
Relaxation
Quality markers:
1,474
Randomisation
description
3 (21%)
Concealment of
allocation
2 (14%)
17
1,357
1 (6%)
0 (0%)
23
1,481
6 (26%)
1 (4%)
Multiple
intervention
Alcohol
reduction
6
413
2 (33%)
4
865
Coffee
0
Calcium
Baseline
comparability a
Blinding of:
12 (86%)
Participant b
16
5
-
Treatment
provider
-
Outcome
assessor
4 (29%)
13 (76%)
-
-
2 (12%)
5 (65%)
-
-
10 (43%)
0 (0%)
5 (83%)
-
-
4 (67%)
1 (33%)
0 (0%)
2 (67%)
-
-
2 (67%)
0
-
-
-
-
-
-
11
414
2 (18%)
1 (9%)
4 (36%)
9 (82%)
9 (82%)
1 (9%)
Magnesium
11
504
1 (9%)
0 (0%)
6 (55%)
9 (82%)
10 (91%)
0 (0%)
Potassium
5
410
3 (60%)
2 (40%)
2 (40%)
3 (60%)
3 (60%)
3 (60%)
Sodium
5
420
0 (0%)
0 (0%)
2 (40%)
0 (0%)
0 (0%)
0 (0%)
Combined salts
2
240
1 (50%)
0 (0%)
2 (100%)
2 (100%)
2 (100%)
0 (0%)
a Confirmation of baseline comparability for parallel trials or of no carryover effect for crossover trials.
b Neither participant nor treatment provider could be blinded to behavioural interventions.
In overview, 98 trials including 7,993 participants were combined to provide principal findings on lifestyle interventions (see Figure 4)
although these were augmented with a number of other trials and reviews. Statistically significant reductions in blood pressure were found, in
the short term for improved diet and exercise, relaxation therapies, and sodium and alcohol reduction. For example, our best estimate is that a
multiple intervention addressing diet and exercise can reduce systolic and diastolic blood pressure in a cohort of patients, on average, by about
5 mmHg. However this estimate is based on a limited number of patients and is uncertain. The 95% confidence interval shows that (19 times
out of 20) the true average reduction may be anywhere between about 2 and 9 mmHg. Individual patients may achieve a greater or lesser
reduction than the average and for a combined diet and exercise intervention the best guess is that about one quarter of patients will achieve a
reduction in systolic blood pressure of at least 10 mmHg.
Hypertension (partial update)
Type of
intervention
Lifestyle interventions
Table 48: Summary characteristics of trials of lifestyle interventions
Hypertension (partial update)
Lifestyle interventions
Figure 4: Overview of lifestyle interventions: effect on systolic and diastolic blood
pressure in randomised trials of patients with raised blood pressure
(≥140/85mmHg)
9.1.1
9.1.2
Most areas featured considerable heterogeneity (i.e. study findings were inconsistent, some
positive and some negative) over and above the variation expected by the normal play of
chance. This heterogeneity tends to limit the strength of recommendation that can be made
about any course of action.
Managing changes in lifestyle
Our systolic (and to a lesser extent our diastolic) blood pressure tends to increase as we grow
older. It is unhelpful to think of a single threshold above which we suddenly have
problematically high blood pressure, although such thresholds can be useful to spur us into
action. A review of our lifestyle helps us to identify changes we can make which may reduce
our blood pressure and thus delay, reduce or remove the need for long term drug therapy as
well as leading to a healthier life. The cumulative trial evidence suggests that individuals who
develop improved habits of regular exercise, sensible diet and relaxation can reduce their
blood pressure. Forming these habits will take determination and support. Health care
professionals can provide advice, encouragement and materials but ultimately may have
limited scope to influence poor dietary habits and inadequate exercise which result in part
from the busy and stressful pace of life and in part from personal choice. Much of the research
evidence for lifestyle change uses regular time spent together in groups for support and
encouragement. Patient and healthcare organisations may be able to help provide patients
with, or point them to local groups which encourage lifestyle change, particularly those
promoting healthy eating and regular exercise.
Diet
Fourteen randomised controlled trials, including 1,474 participants, met the review inclusion
criteria. 18,45,84,138,144,235,262,295,310,406,508,520,545,577,617, 380,495,499,502. Studies most commonly
compared low calorie diets, aimed at overweight patients, with either the patients' usual diet
or with a prescribed 'usual care' diet. In addition, one study compared fish oil capsules with
olive oil capsules (as a control); one study compared diets supplemented with fibre from oats
and wheat; one study compared soy milk with skimmed cows' milk; these studies are
discussed separately498, 158, 510.
The mean age of study participants was 48 years and 62% were male. Only four studies
reported ethnicity and in these about 45% of the participants were white. The median duration
of both treatment and follow-up was 26 weeks, ranging from eight weeks to one year.
166
Hypertension (partial update)
Lifestyle interventions
Randomisation could be confirmed as adequate in only three studies (21%) and concealment
of allocation as adequate in only one (7%). Blinding was confirmed as adequate in six studies
(43%). Treatment and control groups were confirmed as comparable at baseline, with regard
to age, sex and initial blood pressure in 12 studies (86%).
Studies varied in their methods and in definitions of diets prescribed. Some focussed
primarily on low saturated fat, others primarily on weight reduction but in practice there was
considerable overlap of content. Patients were sometimes given advice on other aspects of
lifestyle, such as exercise. Dieticians, nurses or counsellors generally delivered interventions
although in two studies doctors were primarily involved. Two of the studies provided meals
for the participants406,520. Contact between participants and the treatment providers varied
considerably from several times weekly through to occasionally. Crucially, we could identify
no clear system for sub-grouping diet studies: there were too many confounding influences.
There was generally little change in the weight of people in the control groups, whereas
average study losses in dietary intervention groups were between two and nine kilograms.
Average changes in blood pressure, when comparing treatment and control groups, are shown
in Figure 5. Overall, with dietary intervention there was a significant reduction in both
systolic (6.0 mmHg, 95% CI: 3.4 to 8.6) and diastolic (4.8 mmHg, 95%CI: 2.7 to 6.9) blood
pressure. There was no evidence of reporting bias, but significant heterogeneity existed
between studies. Forty percent (95%CI: 33% to 47%) of patients put on diets were likely to
show at least a 10 mmHg reduction in systolic blood pressure. There was no overall
difference in withdrawal when comparing diet and control arms of studies (treatment vs.
control, risk difference 3.6%, 95%CI: −0.1% to 7.2%), although studies varied.
Figure 5: Effect of diet on systolic and diastolic blood pressure in randomised trials of
patients with raised blood pressure
Omission of a study which enrolled abnormally hypertensive patients (mean baseline BP:
170/110 mmHg)508 resulted in a more modest estimate of reduced blood pressure due to diet:
systolic 5.0 mmHg (95% CI: 3.1 to 7.0) and diastolic 3.7 mmHg (95%CI: 2.4 to 5.1).
While soy milk appeared to lower blood pressure when compared to skimmed cows' milk510
and fish oil appeared to lower blood pressure when compared to olive oil135, these findings
were from single small short-term studies and require substantiation by other independent
167
Hypertension (partial update)
Lifestyle interventions
9.1.3
studies. In one small study, supplementing the diet with oats did not appear to lower blood
pressure when compared to wheat158.
The Cochrane Collaboration415 carried out a review which had different inclusion criteria (it
included simple interventions reported up to June 1998, had no restriction on length of follow
up and also used weight loss as an end point) leaving only four studies common to both
reviews. Nevertheless, its conclusions were similar. The recent Canadian guideline reviewed
studies between 1966 and 1996355. Although without a formal meta-analysis, it likewise
concluded that overweight hypertensive patients should be advised to reduce their weight.
Exercise
Seventeen randomised controlled trials of parallel design84,85,162,184,235,246,249,261,341,
18,45,231,391,513,559,575,583,585
including 1,357 participants, met the review inclusion criteria.
Studies most commonly enrolled overweight patients and compared no intervention with a
weekly schedule of three to five sessions of aerobic exercise. One study249 offered advice to
participants whereas all others provided facilities. Three further studies could not be included
because of missing data274,327,604.
The mean age of study participants was 53 years and 58% were male. Only five studies
reported ethnicity and in these about 80% of the participants were white. The median duration
of both intervention and follow-up was 17 weeks, ranging from eight weeks to one year.
Randomisation could be confirmed as adequate in only one study (6%), and concealment of
allocation as adequate in none (0%). Blinding was confirmed as adequate in one study (6%).
Treatment and control groups were confirmed as comparable at baseline, with regard to age,
sex and initial blood pressure in 13 studies (76%).
Overall, patients receiving exercise-promoting interventions achieved a modest reduction in
both systolic (3.1 mmHg, 95%CI: 0.7 to 5.5) and diastolic (1.8 mmHg, 95% CI: 0.2 to 3.5)
blood pressure compared to those in control groups (see Figure 6). There was no evidence of
reporting bias. Significant heterogeneity existed between studies, although there was no
obvious underlying cause for this. There were not enough studies to explore the relative
merits of weight training compared to aerobics or differences between low and medium
intensity aerobics. Thirty-one percent (95% CI: 23% to 38%) of patients receiving exercise
interventions were likely to show at least 10 mmHg reduction in systolic blood pressure.
People in the exercise arms were more likely to withdraw from the studies than those in the
control arms (treatment vs. control, risk difference: 5.9%, 95%CI: 0.1% to 11.1%), although
studies varied.
168
Hypertension (partial update)
Lifestyle interventions
Figure 6: Effect of exercise on systolic and diastolic blood pressure in randomised
trials of patients with raised blood pressure
9.1.4
A recent systematic review of studies of the effect of exercise on blood pressure187 included
seven studies between 1966 and 1995, all with at least 26 weeks follow-up, and including
normotensive and hypertensive participants. The review found exercise had a small and
statistically non-significant effect on blood pressure (−0.7/0.3 mmHg in 4 studies with
hypertensive participants), but noted the poor quality of studies.
The recent Canadian guideline reviewed studies between 1966 and 1997132. Although without
a formal meta-analysis, it reported short term reductions in blood pressure of 5 to 10 mmHg
and recommended 50–60 minutes of moderate intensity exercise three or four times per week.
Relaxation therapies
Twenty-three randomised controlled trials of parallel design, including 1,481 participants, met
the review inclusion criteria. RCTs of relaxation interventions32,33,
31,34,69,95,115,120,142,221,265,276,277,289,304,367,397,477-479,525,533,610,661
. Twelve further trials could not be
included because of missing data128,232,245,345,398,586, 36,80,92,288,418.
The mean age of study participants was 49 years and 62% were male. Only six studies
reported ethnicity and in these about 84% of the participants were white. The median duration
of intervention was 8 weeks, ranging from four weeks to six months; the median duration of
follow-up 17 weeks, ranging from eight weeks to four years, reflecting that studies often
assessed the longer term impact of interventions well after formal therapy had ceased.
Randomisation could be confirmed as adequate in only seven studies (30%), and concealment
of allocation as adequate in only one (4%). Blinding was confirmed as adequate in seven
169
Hypertension (partial update)
Lifestyle interventions
studies (30%). Treatment and control groups were confirmed as comparable at baseline, with
regard to age, sex and initial blood pressure in 16 studies (70%).
The common component in studies was a strategy to promote relaxation although this could
be oriented through education, physical techniques (such as breathing or progressive muscle
relaxation), talk therapies, stress management or some combination. Additionally some
studies used biofeedback, where the participant received auditory or visual information about
their heart rate, peripheral temperature or some other physical marker. There was variation in
content, with individual studies incorporating (for example) forms of cognitive training,
breathing management, meditation, yoga, behavioural contracts, assertiveness training and
anger control techniques. Similarly, delivery varied, being provided by a range of health
professionals, most commonly to groups but in a few studies to individuals. Most treatment
sessions were about an hour in length (varying from 30 to 90 minutes) and were usually
conducted once a week.
Control groups received care varying from no intervention to sham group therapy excluding
components that investigators believed to be the effective aspects of therapy. Some studies
included both types of control groups.
Overall relaxation interventions were associated with statistically significant reductions in
systolic (3.7 mmHg, 95%CI: 1.3 to 6.0) and diastolic (3.5 mmHg, 95%CI: 1.9 to 5.1) blood
pressure (see Figure 7). There was no evidence of reporting bias. However, significant
heterogeneity existed between studies. Analysis of the additional value of biofeedback as a
component of the intervention was inconclusive when comparing studies that did or didn't
include it, or when comparing alternative interventions within trials. Thirty-three percent
(95%CI: 25% to 40%) of patients receiving relaxation therapies were likely to show at least a
10 mmHg reduction in systolic blood pressure in the short term. Based on 12 of the studies,
there was no significant difference in withdrawal when comparing treatment or control arms
of studies (treatment vs. control, risk difference: 3.4%, 95%CI: 0.0% to 6.8%), although
studies varied.
170
Hypertension (partial update)
Lifestyle interventions
Figure 7: Impact of relaxation interventions on blood pressure: findings from
randomised controlled trials
9.1.5
A recent systematic review of studies of the effect of stress reduction on blood pressure187
included seven studies between 1966 and 1995, all with at least 26 weeks follow-up, and
including hypertensive participants. Although the inclusion criteria differed from ours, the
review found a small and statistically non-significant effect on blood pressure (−1.0/−1.1
mmHg) consistent with longer follow-up studies reported here. The review similarly found
considerable heterogeneity between studies.
The recent Canadian guideline reviewed studies between 1966 and 1997550. It concluded that
multifaceted interventions to reduce stress were more likely to be effective than single
component therapies and favoured the use of cognitive behavioural therapy, based on the
findings of three meta-analyses192,293,366. For hypertensive patients in whom stress appears to
be an important issue, they recommended that stress management including individualized
cognitive behavioural therapy may be appropriate.
Multiple lifestyle interventions
Six randomised controlled trials, including 413 participants, met the review inclusion criteria.
RCTs of multifaceted interventions45,47,84,294,337,337,408,599. Three of the studies essentially
provided a therapeutic intervention combining group exercise and diet strategies similar to the
lifestyle interventions found in the previous sections45,47,84,337, 599; one study also included
relaxation and restriction of intake of common salt337; one study combined a weight loss diet,
relaxation and salt restriction294; and one study combined a weight loss diet, exercise and salt
restriction408. A further trial, which delivered a health education package to a British
population with angina, did not meet our inclusion criteria for blood pressure and so was
171
Hypertension (partial update)
Lifestyle interventions
excluded from the meta-analysis and is considered separately146. Three further trials could not
be included because of missing data274,309,334.
The mean age of participants was 52 years, 66% were male and the median follow-up of
studies was six months. Five studies reported ethnicity and in these about 75% of the
participants were white.
Randomisation was confirmed as adequate in only two studies (33%). Concealment of
allocation was inadequate or unclear in all six studies. Blinding was confirmed as adequate in
four studies (67%). Treatment and control groups were confirmed as comparable at baseline,
with regard to age, sex and initial blood pressure in five studies (83%).
Overall, multifaceted interventions caused a modest reduction in both systolic (5.5, 95%CI:
2.3 to 8.8) and diastolic (4.5 mmHg, 95% CI: 2.0 to 6.9) blood pressure (see Figure 8).
However heterogeneity existed between studies: the study of Jacob (1985) did not
demonstrate a reduction in blood pressure. Twenty-six percent (95%CI: 2% to 49%) of
patients receiving combined interventions were likely to show at least a 10 mmHg reduction
in systolic blood pressure. Data from five studies found no statistically significant difference
in withdrawal from treatment and control groups (treatment versus control, risk difference:
4.9%, 95%CI: −2.6% to 12.4%).
Figure 8: Impact of combined lifestyle interventions on blood pressure: findings from
randomised controlled trials
It was not possible to assess from the available data whether the effects of diet and exercise
were additive or whether the combination was no better than either diet or exercise on its
own.
The large British health promotion study, of 688 participants, lasted longer (two years) and
was of older people (mean age 63 years) than the therapeutic studies. It did not show any
reduction in blood pressure in response to health advice, but nevertheless reported fewer
deaths among those receiving advice (29 in control group and 13 in treatment group),
providing a relative reduction in mortality of 55%, an absolute reduction in mortality of 4.6%
(95%CI: 1.0% to 8.4%) or a Number Needed to Treat of 22 to prevent a death during two
years of follow-up. Patients in this trial, suffering from angina, were at higher risk than most
other patients enrolled in lifestyle trials, leading to greater levels of morbidity and mortality.
However, the benefit of health promotion in this trial does not appear mediated by reduced
blood pressure or any other obvious prognostic marker (smoking, cholesterol or body mass
index), and thus needs confirmation from further research.
A recent systematic review of studies of multiple interventions for preventing coronary heart
disease; included nine studies of normotensive and hypertensive participants, published
172
Hypertension (partial update)
Lifestyle interventions
9.1.6
between 1966 and 1995, and with at least 26 weeks follow-up186. The review found an overall
reduction of 4.2/2.7mmHg, but no significant reductions in morbidity and mortality in studies
not including drug interventions.
Alcohol
The epidemiological link between alcohol consumption, blood pressure, cardiovascular
disease and all-cause mortality has been studied extensively181,263,497,596. While moderate
consumption may do no harm, the literature consistently finds that the move from moderate to
excessive drinking (men: more than 21 units/week; women: more than 14 units/week) is
associated both with raised blood pressure and a poorer prognosis. (Approximately: one halfpint of beer, glass of wine or a single measure of spirits equals one unit of alcohol or one
standard drink and contains 8g or 10ml of alcohol287).
Three randomised controlled trials, including 397 participants, met the review inclusion
criteria and examined the effect of changes in alcohol consumption on blood
pressure148,382,502. Interventions varied in their content but commonly featured a number of
visits to a health care practitioner for advice on reducing intake of alcohol. At baseline,
patients typically reported drinking 300 to 600 ml of alcohol, or 30–60 standard drinks, per
week. Although alcoholism was not formally defined, very heavy drinkers were commonly
excluded. A further cluster randomized trial with 93 participants was identified and included
in a secondary analysis348.
The mean age of study participants was 53 years; in the two studies that provided the details
all participants were male and three quarters were white. The PATHS study148, with 6 months
treatment duration, two year follow-up and 59% of patients, differed in scale from the two
other shorter and smaller trials.
Randomisation could be confirmed as adequate only in the PATHS study, and concealment of
allocation as adequate in none. Blinding was confirmed as adequate in two studies. Treatment
and control groups were confirmed as comparable at baseline, with regard to age, sex and
initial blood pressure in all three studies, with the possible exception of PATHS which did not
report the proportions of men and women in the treatment and control groups. No studies
were designed to assess the impact of alcohol reduction on cardiovascular endpoints.
Overall, interventions to reduce alcohol consumption caused small but statistically significant
reductions in both systolic (3.4 mmHg, 95%CI: 0.9 to 6.0) and diastolic (3.4 mmHg, 95%CI:
1.5 to 5.4) blood pressure. Thirty percent (95%CI: 21% to 39%) of patients receiving a
structured intervention to reduce alcohol consumption were likely to achieve a reduction of at
least 10 mmHg in systolic blood pressure. No harmful effects of intervention were reported in
these trials; withdrawal rates were reported in only one small trial. Inclusion of the single
cluster randomized study did not alter qualitatively the summary reduction in systolic (3.7
mmHg, 95% CI: 1.3 to 6.1) or diastolic (3.2 mmHg, 95%CI: 1.4 to 5.0) blood pressure, (see
Figure 9).
Figure 9: Impact of alcohol reduction on blood pressure: findings from randomised
controlled trials
173
Hypertension (partial update)
Lifestyle interventions
9.1.7
9.1.8
The recent Canadian guideline reviewed studies between 1966 and 1996113. Although without
a formal meta-analysis, it recommended that alcohol consumption be limited in patients with
hypertension to two or fewer standard drinks per day, with consumption not exceeding 14
standard drinks per week for men and nine standard drinks per week for women.
For recommendations on preventing the development of hazardous and harmful drinking, see
NICE Public Health guidance 24 (http://guidance.nice.org.uk/PH24).
Coffee
Although coffee is a complex beverage containing many chemicals, only the effect of caffeine
has been studied extensively516. According to personal taste and type of coffee, the amount of
caffeine varies, but typically coffee contains 60 to 120 mg per 150ml cup. This can be
compared with tea (20 to 40 mg per 150ml cup) and cola drinks (30 to 50 mg per 330ml
can)444, 130.
Caffeine consumption has long being associated with raised blood pressure and can
demonstrate a dose-related increase of 5–15 mmHg systolic and 5–10 mmHg diastolic for
several hours following consumption. The most likely mode of action of caffeine is as an
adenosine receptor antagonist, which results in vasoconstriction and raises blood pressure.
The half life of caffeine in the body is typically about five hours297.
We identified no randomised controlled trials examining the impact of coffee or caffeine
intake on patients with hypertension, which provided at least eight weeks follow-up. A
published systematic review included normotensive as well as hypertensive participants, and
shorter durations of follow-up299. Eleven trials with a total of 522 participants and a median
duration of eight weeks (range 2 to 11 weeks) were included. Control groups drank a median
of five caffeinated cups of coffee a day, with treatment groups receiving no, or decaffeinated,
coffee. The reported overall effect of coffee was an increase in systolic (2.4 mmHg, 95%CI:
1.0 to 3.7) and diastolic (1.2 mmHg, 95%CI: 0.4 to 2.1) blood pressure.
Identifying the influence of coffee upon blood pressure, or identifying groups at particular
risk, is problematic in the presence of confounding factors such as age, lifestyle, and
cardiovascular disease. The small sample sizes and durations of existing trials do not provide
an adequate evidence base to infer the long term effects of routine caffeine consumption.
Reducing sodium (salt) intake
Practical steps to reduce sodium intake include choosing low-salt foods (e.g. choosing fresh
fruits and vegetables and avoiding processed foods) and reducing or substituting its use in
cooking and seasoning. Much dietary salt comes from processed foods whose content should
be labelled helping to monitor intake.
Five randomised controlled trials (four of parallel design125,212,311,544, one of crossover
design10,11), examining the effect of sodium reduction on blood pressure, met the review
inclusion criteria and included 420 patients. The findings of one Italian trial in young adults
are considered separately141. A further trial could not be included because of missing data395.
The mean age of study participants was 52 years and 81% were male. The ethnicity of
participants was not reported in any of the studies. The median duration of both intervention
and follow-up was 12 weeks.
One trial (17%) was double-blinded; blinding could not be confirmed in any of the other
studies. Randomisation and concealment of allocation could not be confirmed to be adequate
in any of the studies. Treatment and control groups were confirmed as comparable at baseline,
with regard to age, sex and initial blood pressure in 2 studies of parallel design (40%); the
crossover study did not report on carryover effects.
The studies advised participants to change their diet so as to restrict their sodium intake to
below 70–100 mmol/day (4.2 – 6.0g of salt). The Scientific Advisory Committee on Nutrition
target for all adults is 6 grams/day532 and NICE public health guidance on the prevention of
174
Hypertension (partial update)
Lifestyle interventions
cardiovascular diseases recommends people aim for a maximum intake of 6 grams per day per
adult by 2015 and 3 grams by 2025.
Average changes in blood pressure, when comparing treatment and control groups, are shown
in Figure 10. Sodium reduction was associated with a statistically significant reductions in
systolic (3.4 mmHg, 95%CI: 2.3 to 4.5) and diastolic (2.2 mmHg, 95%CI: 1.5 to 3.0) blood
pressure. Twenty-three percent (95%CI: 17% to 30%) of patients who reduced their salt
intake were likely to show at least a 10 mmHg reduction in systolic blood pressure. Based on
two studies, there was no difference in withdrawal when comparing treatment and control
arms of studies (treatment versus control, risk difference: −0.6%, 95%CI: −6.5% to 5.4%).
Figure 10:
Impact of sodium reduction on blood pressure: findings from
randomised controlled trials
9.1.9
One Italian trial enrolled young, borderline hypertensive participants, aged 16–31 years. This
trial found a dramatic reduction in systolic (18.4 mmHg, 95%CI: 10.1 to 26.7) blood pressure.
The trial was poorly described and it is unclear whether the reduction in systolic blood
pressure is due solely to the intervention. The authors note that the benefit was found mostly
in participants less than 20 years of age. The inclusion of the trial in the meta-analysis
increased the average benefit of salt reduction on systolic blood pressure (7.1 mmHg, 95%CI:
2.9 to 11.3), but introduced considerable statistical heterogeneity (Q: p=0.007).
Two recent systematic reviews have evaluated advice to reduce salt intake in normotensive
and hypertensive adults, in trials with at least 6 months follow-up187,279. The inclusion criteria
used in these reviews differ from ours, notably they included studies where the dose of
antihypertensive drugs was allowed to vary. Regardless, both reviews found statistically
significant reductions in blood pressure in studies with hypertensive participants, of 2.5/1.2
(up to one year follow-up) and 1.1/0.6 (one to six years follow-up)279 and 2.9/2.1 mmHg187,
suggesting that reductions in blood pressure tend to diminish over time.
The recent Canadian guideline220, citing a previous systematic review, concluded that sodium
restriction in adults over 44 years of age resulted in a reduction in blood pressure of 6.3/2.2
mmHg per 100 mmol/day reduction in sodium. Recommendations were made for clinicians to
determine salt intake by interview; aim for a target range of 90–130 mmol per day (3–7 grams
per day); provide advice on choosing low-salt foods (e.g. choosing fresh fruits and vegetables
and avoiding pre-prepared foods) and reduce usage in cooking and seasoning.
Calcium supplements
Eleven randomised controlled trials (three of parallel design242,378,442, eight of crossover
design227,318,396,571,581,584,627,660), examining the effect of calcium supplementation on blood
pressure, met the review inclusion criteria and included 414 patients. Another trial, carried out
in patients who were undergoing dialysis, was excluded after consideration of their unusual
175
Hypertension (partial update)
Lifestyle interventions
calcium metabolism but its details are tabulated487. A further trial could not be included
because of missing data414.
The mean age of study participants was 45 years and 68% were male. Only four studies
reported ethnicity and in these 46% of the participants were white. The median duration of
both intervention and follow-up was eight weeks.
Randomisation could be confirmed as adequate in only two studies (18%) and concealment of
allocation as adequate in only one (9%); nine studies (82%) studies were double-blinded
treatment and control groups were confirmed as comparable at baseline, with regard to age,
sex and initial blood pressure in one study (33%) of parallel design; three studies (37%) of
crossover design confirmed no carryover effect.
The intervention was provided as a simple oral supplement taken several times a day.
Average changes in blood pressure, when comparing treatment and control groups, are shown
in Figure 11. Calcium supplementation was associated with a small reduction in systolic
blood pressure 2.3 mmHg, 95%CI: 0.3 to 4.4) which was statistically significant but not
robust to minor changes in the reported blood pressure of the participants, and no difference
in diastolic blood pressure (−0.8 mmHg, 95%CI: −2.1 to 0.6). No harmful effects of
intervention were reported in these trials; withdrawal rates were on average around 10% in
both treatment and control groups. The trials were unable to identify sub-groups of patients
that might benefit from calcium.
Figure 11:
Impact of calcium supplementation on blood pressure: findings from
randomised controlled trials
9.1.10
Magnesium supplements
Eleven randomised controlled trials (nine of parallel design215,270,365, 91,443,475,621,646,659] 2 of
crossover design [317,645), examining the effect of magnesium supplementation on blood
pressure, met the review inclusion criteria and included 504 patients.
The mean age of study participants was 55 years and 44% were male. Only two studies
reported ethnicity and in these 11% of the participants were white. The median duration of
both intervention and follow-up was 12 weeks.
Ten studies (91%) studies were single or double blinded. Randomisation and concealment of
allocation were confirmed to be adequate in one study (9%) and no studies respectively.
Treatment and control groups were confirmed as comparable at baseline, with regard to age,
176
Hypertension (partial update)
Lifestyle interventions
sex and initial blood pressure in six studies (67%) of parallel design; neither of the studies of
crossover design reported on carryover effects.
The intervention was provided as a simple oral supplement taken several times a day.
Average changes in blood pressure, when comparing treatment and control groups, are shown
in Figure 12. Magnesium supplementation was associated with little change in systolic (−1.0
mmHg, 95%CI: −4.1 to 2.1) but a statistically significant reduction in diastolic (−2.1 mmHg,
95%CI: −3.5 to −0.7) blood pressure. No harmful effects of intervention were reported in
these trials; withdrawal rates were reported in only eight studies, where these were on average
around 7% in both treatment and control groups. The trials were unable to identify sub-groups
of patients that might benefit from magnesium.
Figure 12:
Impact of magnesium supplementation on blood pressure: findings from
randomised controlled trials
9.1.11
Potassium supplementation
Five randomised controlled trials (four of parallel design107,543,543, 578, one of crossover
design470), examining the effect of potassium supplementation on blood pressure, met the
review inclusion criteria and included 410 patients. The findings of one African trial are
considered separately455. A further trial could not be included because of missing data149.
The mean age of study participants was 51 years and 76% were male. Only one study
reported ethnicity and in this 86% of the participants were white. The median duration of both
intervention and follow-up was 12 weeks.
Two studies were triple blinded, two were assessment blinded and one was unclear.
Randomisation and concealment of allocation were confirmed to be adequate in one (20%)
and two (40%) studies respectively. Treatment and control groups were confirmed as
comparable at baseline, with regard to age, sex and initial blood pressure in two studies (50%)
of parallel design; the crossover study did not report on carryover effects.
The intervention was provided as a simple oral supplement taken several times a day in all but
one trial, where dietary advice was provided to increase intake of foods rich in potassium125.
Average changes in blood pressure, when comparing treatment and control groups, are shown
in Figure 13. Potassium supplementation was not associated with any significant change in
systolic (−3.5 mmHg, 95%CI: −7.9 to 0.9) or diastolic (−0.7 mmHg, 95%CI: −4.9 to 3.6)
blood pressure. The findings of the studies were heterogeneous and there are no obvious
reasons for this that can be deduced from the limited available evidence. No harmful effects
177
Hypertension (partial update)
Lifestyle interventions
of intervention were reported in these trials; average withdrawal rates of 6–8% were similar in
both treatment and control groups.
Figure 13:
Impact of potassium supplementation on blood pressure: findings from
randomised controlled trials
9.1.12
One trial, which enrolled treatment naïve and hypertensive Kenyan participants (DBP 90–109
mmHg and SBP>160 mmHg) reported an average reduction of 39/17 mmHg. Although the
effect of various salts upon certain ethnic groups is known to vary, a reduction of this
magnitude exceeds our understanding and requires confirmation from further independent
research.
A meta-analysis by Whelton and colleagues found that oral potassium supplementation was
associated with a significant reduction in both systolic blood pressure and diastolic blood
pressure633, based on 12 trials in normotensive people and 21 in hypertensive people, with a
duration ranging from four days to three years (median five weeks). The review found that the
blood pressure lowering effect was greater in hypertensive than normotensive people,
although the statistical significance of findings in the hypertensive subgroup is not reported.
The review also found that the effect was more pronounced in people eating a diet high in
sodium chloride (common salt) and therefore recommended potassium supplementation for
both prevention and treatment of hypertension, especially in people unable to reduce their
intake of sodium.
In contrast, our restriction to trials of at least 8 weeks duration, enrolling only hypertensive
patients, resulted in inclusion of only 5 trials with a median duration of 12 weeks and found
that the blood pressure lowering effect of oral potassium supplementation was not statistically
significant. The group concluded that there is not sufficient relevant evidence to recommend
oral potassium supplementation for hypertension.
Combined salt supplements
Two randomised controlled trials studied combinations of the potassium, magnesium, sodium
and calcium salts considered individually in previous sections.
One study used paired supplements comparing two of calcium, potassium and magnesium
with placebo519. None of the combined supplements reduced blood pressure when compared
with placebo (see Figure 14). This was consistent with the findings for the individual
supplements.
178
Hypertension (partial update)
Lifestyle interventions
Figure 14:
Impact of combined supplements on blood pressure: findings from
randomised controlled trials
9.1.13
A second study compared a mineral (reduced sodium) salt containing sodium, potassium and
magnesium with common sodium table salt. The mineral salt was used in prepared food as
well as for seasoning229. The reduction of blood pressure by about 5/4 mmHg consistent with
that found with strategies to reduce sodium salt intake.
The recent Canadian guideline reviewed studies between 1966 and 1996108. Although without
a formal meta-analysis, it recommended against supplementing calcium, magnesium or
potassium intake amongst hypertensive participants above the recommended normal daily
levels.
Drug therapy versus lifestyle change
Five small randomised controlled trials enrolling 233 patients directly compared the effects of
lifestyle interventions and drugs for the treatment of mild to moderate hypertension. Goldstein
et al 232, Murugesan et al 418, Kostis et al 337, MacMahon et al 380, 381, Koopman et al 333. An
additional quasi-randomised trial, which allocated participants to treatments on the basis of
their birth date rather than at random, was also considered (Berglund et al72).
All trials were small (between 38 and 66 participants), of short duration (between eight and 52
weeks) and were not designed to assess cardiovascular endpoints. Randomisation and
concealment of allocation were either inadequate or not clearly reported in all trials. The
outcome assessor was blinded to the treatment status of the participants in three trials333,337,380;
blinding was not reported in two trials232,418, and there was no blinding in one trial72. One trial
was poorly reported and did not state the total number of participants418. In two trials the
confidence intervals on the effects of treatment could not be estimated, as either the numbers
in each treatment group418 or the standard error of the treatment effects were not reported232.
The populations studied in the trials differed in: (i) age – participants in one trial333 were
older, which probably accounted for their higher baseline blood pressure compared to
participants in the other trials; (ii) treatment status at the point of recruitment – participants
were currently untreated or treatment naïve in four trials72,232,333,380, currently treated in one
trial337, or treatment status was not reported418.
The trials compared different drugs with different lifestyle interventions. Typically either a
diuretic or a beta-blocker was the class of drug used, although one trial allowed a choice of
drugs. Four trials used a low calorie diet: one used diet alone; one combined a low calorie
intake with a low sodium and high potassium diet; one used a multiple intervention
combining weight loss, a low calorie and low sodium diet, exercise, and relaxation and one
combined weight reduction with restricted sodium and alcohol intake. Two trials had
relaxation interventions: one considered two separate relaxation interventions (biofeedback
and muscular relaxation/breathing exercises); the other used yoga.
179
Hypertension (partial update)
Lifestyle interventions
Five trials reported comparable blood pressure at baseline in both treatment groups and for
one trial this was unclear. Within each study, findings for systolic and diastolic blood pressure
were similar.
Trials comparing diet with drugs provided conflicting evidence (see Figure 15). In the trial of
older participants333 who had not received treatment before and had a high baseline blood
pressure, drug treatment appears more effective than diet in lowering blood pressure, whereas
in a trial of younger participants381 who were currently untreated and had a lower initial blood
pressure, diet appears significantly more effective than drug treatment in lowering blood
pressure. The one trial337 comparing multiple lifestyle interventions with drugs found both
treatments had similar effects on lowering blood pressure. Two trials found drugs to be more
effective than relaxation although the confidence intervals on the treatment effects could not
be determined418.
Figure 15:
Comparison of lifestyle and drug interventions: findings from
randomised controlled trials
Participants receiving dietary interventions improved their total cholesterol profiles in all four
trials compared to participants receiving drugs. Cholesterol levels were not reported in either
relaxation trial. Although it was a post hoc exercise, we combined cholesterol reductions
found in the dietary trials by imputing missing standard deviations. Using a random effects
model, the average reduction in cholesterol was 0.52 mmol/l (95% CI −0.34 to −0.7).
Withdrawals were reported in five trials: rates of withdrawal were similar for lifestyle and
drug treatments.
The current evidence cannot determine whether a lifestyle intervention is generally better than
drug treatment for reducing blood pressure. Although cholesterol levels were not a
prespecified outcome, it was observed that, in all four trials with diet interventions, diets were
better than antihypertensive drugs at reducing cholesterol. As reduced cholesterol levels are
likely to lower the risk of cardiovascular morbidity or mortality irrespective of any change in
blood pressure643, a healthier diet may reduce, delay or remove the need for long-term drug
therapy in some patients. Thus it seems important that patients are encouraged to try lifestyle
changes before proceeding to or increasing drug therapy.
9.1.14
Smoking cessation
A review of the health consequences of smoking and benefit of smoking cessation is not
included in this guideline, since there is no direct link to raised blood pressure. However
smoking reduces life expectancy and is associated with poor cardiovascular and pulmonary
outcomes179,180,357,410,488,648. The NHS website www.smokefree.nhs.uk has facts and
information about giving up smoking.
180
Hypertension (partial update)
Pharmacological interventions
9.1.15
Refer to NICE’s public health guidance on smoking cessation services in primary care,
pharmacies, local authorities and workplaces, particularly for manual working groups,
pregnant women and hard to reach communities for more information
(www.guidance.nice.org.uk/PH10).
Recommendations
For NICE guidance on the prevention of obesity and cardiovascular disease see ‘Obesity’
(NICE clinical guideline 43, 2006) and ‘Prevention of cardiovascular disease at population
level’ (NICE public health guidance 25, 2010).
31. Lifestyle advice should be offered initially and then periodically to people undergoing
assessment or treatment for hypertension. [2004]
32. Ascertain people’s diet and exercise patterns because a healthy diet and regular exercise
can reduce blood pressure. Offer appropriate guidance and written or audiovisual materials
to promote lifestyle changes. [2004]
33. Relaxation therapies can reduce blood pressure and people may wish to pursue these as
part of their treatment. However, routine provision by primary care teams is not currently
recommended. [2004]
34. Ascertain people’s alcohol consumption and encourage a reduced intake if they drink
excessively, because this can reduce blood pressure and has broader health benefits. [2004]
35. Discourage excessive consumption of coffee and other caffeine-rich products.
36. Encourage people to keep their dietary sodium intake low, either by reducing or
substituting sodium salt, as this can reduce blood pressure.[2004]
37. Do not offer calcium, magnesium or potassium supplements as a method for reducing
blood pressure. [2004]
38. The best current evidence does not show that combinations of potassium, magnesium and
calcium supplements reduce blood pressure. [2004]
39. Offer advice and help to smokers to stop smoking. [2004]
40. A common aspect of studies for motivating lifestyle change is the use of group working.
Inform people about local initiatives by, for example, healthcare teams or patient
organisations that provide support and promote healthy lifestyle change. [2004]
10
Pharmacological interventions
In most hypertensive patients, pharmacological intervention becomes necessary if blood
pressure lowering is to be substantial and sustainable. Published epidemiological studies and
trials together conclusively demonstrate that a sustained reduction in blood pressure by drugs
reduces the incidence of stroke, coronary heart disease, heart failure and mortality. The size of
benefit in any period (for example the next 10 years) generally depends on an individual's
overall cardiovascular risk. 135,379 For an individual at any age, the greater the cardiovascular
risk the greater the potential to benefit from treatment.
The Department of Heath National Service Framework for Coronary Heart Disease standards
3 and 4 relate to patients at risk of cardiovascular disease. 'General practitioners and primary
care teams should identify all people with established cardiovascular disease and offer them
comprehensive advice and appropriate treatment to reduce their risks'. 'General practitioners
and primary health care teams should identify all people at significant risk of cardiovascular
disease but who have not developed symptoms and offer them appropriate advice and
181
Hypertension (partial update)
Pharmacological interventions
treatment to reduce their risks.' Similarly, the Welsh National Service Framework for
Coronary Heart Disease states, 'Everyone at high risk of developing coronary heart disease ...
should have access to a multifactorial risk assessment and be offered an appropriate
treatment plan'.
Based on the findings of trials, a range of drugs (some blood pressure lowering) are offered to
patients with existing coronary heart disease. These patients are the subject of a previously
published national guideline. 440 The recommendations include the use of aspirin, betablockers, statins and ACEi. Once patients are optimally treated to prevent further disease,
persistent hypertension should be managed adapting the recommendations from this
document.
Trials treating raised blood pressure, and described in this guideline, include patients both
with and without cardiovascular disease and thus are relevant to the management of raised
blood pressure in all of these patients after any disease specific care has been delivered.
Drugs for raised blood pressure are prescribed alone or in combination, and aim to control
blood pressure while minimising side effects or toxicity. How the drugs work is not always
fully understood. A brief summary of drugs used for essential hypertension is provided in
Table 49; further information can be found in the British National Formulary. 306 Drugs for
hypertension rarely have serious side-effects when appropriately initiated and adequately
monitored.
Table 49: Outline of drugs used for essential hypertension
Commonly used Classes of Antihypertensive Drug Therapies in the United Kingdom
(This is intended as a guide and reference to the product label and British National Formulary is
recommended for detailed prescribing information)
Common generic
Mode of action
Duration of
Usage notes
Class
names
action
bendroflumethiazide,
Vasodilation and
Commonly
Can cause gout and
Thiazide
moderate
diuresis
once
daily
hypokalaemia and rarely
diuretics
hydrochlorthiazide
(increased
morning use
hyponatraemia.
excretion of
Can increase the risk of
sodium,
developing type 2 diabetes
potassium and
water).
Chlortalidone,
Vasodilation and
Commonly
Can cause gout and
Thiazide –
moderate
diuresis
once
daily
hypokalaemia and rarely
like diuretics
indapamide
(increased
morning use
hyponatraemia.
excretion of
Can increase the risk of
sodium,
developing type 2 diabetes
potassium and
water).
Potassiumsparing
diuretics
Spironolactone
amiloride
Vasodilation and
moderate diuresis
(increased
excretion of
sodium,
potassium and
water).
182
Once or twice
daily
Used for resistant
hypertension.
Spironolactone can cause
gynaecomastia in males.
Not to be used with
potassium supplements.
Can cause hyperkalaemia,
especially in patients with
impaired renal function.
Should be avoided in
primary care patients with
a baseline potassium
>4.5mmol/L and used with
caution in people with
renal impairment. Careful
Hypertension (partial update)
Pharmacological interventions
Commonly used Classes of Antihypertensive Drug Therapies in the United Kingdom
(This is intended as a guide and reference to the product label and British National Formulary is
recommended for detailed prescribing information)
monitoring of potassium
and renal function is
required..
Suppress plasma
Vary by drug
Not recommended as a
Beta-blockers atenolol, bisoprolol,
metoprolol,
renin production.
from once to
preferred therapy for
Negative
several times
hypertension. Can be
propranolol, sotalol
inotropic and
considered for resistant
daily
chrontropic
hypertension or as an
effects on the
initial therapy for women
heart. Betaof child bearing potential.
blockers with
Also used for patients with
alpha receptor
angina, post myocardial
activity also
infarction and chronic
produce
heart failure.
Contraindicated with
vasodilatation
asthma, heart-block or in
combination with a ratelimiting calcium-channel
blocker.
Reported side-effects
include lethargy,
depression and sleep
disturbance.
Increased risk of type 2
diabetes, especially when
combined with thiazide or
thiazide-like diuretics.
Calciumchannel
blockers
Angiotensin
converting
enzyme
(ACEi)
inhibitors
'dihydropyridines'
amlodipine,
felodipine, lacidipine
nifedipine.
Vasodilatation
and natiuresis
vasculature.
'rate-limiting CCBs'
diltiazem, verapamil
Heart rate
slowing,
vasodilatation and
natiuresis
captopril, enalapril,
lisinopril, perindopril,
ramipril, trandolapril
Inhibition of
angiotensin
coverting enzyme
and reduced
angiotensin II
production.
183
Vary by drug
from once to
twice daily.
Note only
modified
release
formulation of
nifedipine
should be used
to treat
hypertension
Once or twice
daily for
longer acting
forms
Vary by drug
from once to
several times
daily
Reported side-effects
include initial headaches,
palpitations, facial flushing
and ankle swelling.
Caution against use in
heart failure or use with a
beta-blocker.
Reported side-effects
similar to dihydropyridines
but also include
constipation (verapamil)
and skin rashes (diltiazem)
Contraindicated in
pregnancy.
.Careful monitoring of
potassium levels and renal
function required in people
with renal impairment.
Adverse effects include a
persistent dry cough, rash
and loss of taste. Rarely
angioedema which is more
Hypertension (partial update)
Pharmacological interventions
Commonly used Classes of Antihypertensive Drug Therapies in the United Kingdom
(This is intended as a guide and reference to the product label and British National Formulary is
recommended for detailed prescribing information)
common in black people of
African or Caribean origin
candesartan,
Selective
Once daily
Contraindicated in
Angiotensin
irbesartan, losartan,
inhibition of the
pregnancy.
receptor
olmersartan, valsartan, angiotensin AT-1
blockers
Careful monitoring of
telmisartan
receptor.
(ARBs)
potassium levels and renal
function required in people
with renal impairment.
Generally well tolerated
and unlike ACEi, do not
cause cough
Alpha
receptor
blockers
doxazosin, prazosin,
terazosin
Antagonists of the
Alpha 1 receptor.
Vary by drug
from once to
several times
daily
Consider for the treatment
of resistant hypertension.
Beneficial side-effect on
blood lipid profile.
May also be considered for
men with symptoms of
prostatic outflow
obstruction. Caution in
women in whom they may
cause or worsen symptoms
of stress incontinence.
Contraindications, cautions
and side-effects vary by
drug.
Most common side-effects:
initial dizziness, postural
hypotension, headache,
flushing, nasal congestion,
fluid retention, ankle
swelling and tachycardia.
10.1 2004 guidance: pharmacological interventions
10.1.1
Placebo controlled trials
An overview of key design characteristics of the 20 placebo controlled trials identified is
shown in Table 50 (22 trials are tabulated since two trials had additional treatment arms).
Seldom was the method of randomisation or steps to conceal allocation from investigators or
patients adequately described, although this reflects contemporary standards of reporting.
Patients, clinicians and assessors were commonly blind to the treatment received although
individual trials varied.
Table 50: Summary of characteristics of placebo controlled trials
Number of studies
Thiazides
(High
Dose)
Thiazides
(Low Dose)
Beta
Blockers
Ca
Channel
Blockers
ACEi
Angiotensin
Receptor
Blockers
7
5
7
1
1
1
2 (29%)
0 (0%)
3 (43%)
1 (100%)
1 (100%)
1 (100%)
Quality markers:
Randomisation
description
184
Hypertension (partial update)
Pharmacological interventions
Thiazides
(High
Dose)
Thiazides
(Low Dose)
Beta
Blockers
Ca
Channel
Blockers
ACEi
Angiotensin
Receptor
Blockers
0 (0%)
3 (60%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
Participant
6 (86%)
5 (100%)
6 (86%)
1 (100%)
1 (100%)
1 (100%)
Treatment
provider
4 (57%)
4 (80%)
4 (57%)
1 (100%)
1 (100%)
1 (100%)
Outcome assessor
5 (71%)
4 (80%)
6 (86%)
1 (100%)
0 (0%)
1 (100%)
Baseline
comparability
5 (71%)
5 (100%)
6 (86%)
1 (100%)
1 (100%)
1 (100%)
Concealment of
allocation
Blinding:
Many trials used stepped care regimes aiming to reduce blood pressure to a specified target by
adding other drugs to first line therapy: most of these trials provided matching placebo
stepped care to the control group (ANBPS, VA-NHLBI, EWPHE, SHEP, SHEP-P, SYSTEUR), but some provided no stepped care in the control group (MRC, MRC-O) and some
provided the same active antihypertensive drugs as stepped care to both the active treatment
and the control groups (IPPPSH, SCOPE).
10.1.1.1
Thiazide-type diuretics
Thiazide-type diuretics (thiazides for short) include drugs classified by the British National
Formulary (BNF) as a thiazide or thiazide like diuretic. Twelve trials were identified that met
the review inclusion criteria, see Table 51. Seven trials, with 19,933 participants, starting
from as early as 1964, studied high dose thiazides which are no longer used because of the
risk of complications due to changed plasma potassium, uric acid, glucose, and lipids, with
little additional blood pressure lowering effect compared to low dose thiazides. 26 The mean
age of participants was 51, 59% were male and the mean duration of follow-up was 4.0 years.
Five trials with 15,086 participants, starting between 1975 and 1989, studied low dose
thiazides. Patients had a mean age of 67 years, 53% were male and the mean duration of
follow-up was 4.0 years. Only two studies reported ethnicity and in these 86% of participants
were Caucasian. 'Low dose' is taken pragmatically to mean the doses used in 'low dose' trials
and now normally recommended by the BNF. Although the dichotomisation of low and high
dose used in this guideline for placebo and head-to-head trials is the one commonly used by
reviewers, individual thiazides may sometimes be used at even lower doses.
The underlying risk of disease in patients was proxied by the mortality rate in the control
groups of the trials. HSCSG and PATS enrolled patients following a stroke, but it is
interesting to note the apparent role of age. The underlying risk in PATS is similar to three
other low dose thiazide trials in which patients are, on average, ten years older. It is unclear
why the underlying risk in the EWPHE trial is so high, but this may be due to inclusion of
patients with coronary heart disease. Two trials, SHEP and SHEP-P exclusively enrolled
patients with isolated systolic hypertension (SBP 160–219 mmHg and DBP less than 90
mmHg).
185
ANBPS4
Chlorothiazide
high3
500–
1000
Australia
4.0
1973
30–69
50
Baseline
BP,
mmHg
157/101
HSCSG2
Methychlothiazide
high
10
US
2.1
1966
<75
59
167/100
452
53
MRC402
Bendroflumethiazide
high
10
UK
4.9
1977
35–64
52
161/98
12,951
7
Chlorothiazide
high
50
Norway
5.5
1972
40–49
45
156/97
785
4
Chlorothiazide
high
1000
US
>7
1965
<55
44
147/99
422
3
Oslo
Thiazide1
356
USPHS
VAII
548
1
Dose
category
Dose,
mg
Country
Followup, yrs
Start
year
Age in years
Range
Mean
Number
enrolled
Baseline
Risk2
3,931
5
186
Chlorothiazide
high
100
US
3.2
1964
-
51
164/104
380
39
VA-NHLBI3
Chlorthalidone
high
US
1.5
1978
21–50
38
-
1,012
0
EWPHE6,42,453
Hydrochlorothiazide
low3
50–
100
25–50
Europe
4.7
1975
60+
72
183/101
840
77
MRC-O15
Hydrochlorothiazide
low
25–50
UK
5.8
1982
65–74
70
185/91
3,294
24
PATS
20
Indapamide
low
2.5
China
2.0
1989
-
60
154/93
5,665
28
281,484,485
Chlorthalidone
low
25–50
US
2.8
1981
60+
72
172/75
551
23
13,483,536,606
Chlorthalidone
low
SHEP-P
SHEP
12.5–
US
4.5
1985
60+
72
170/77
4,736
23
25
All trials featured co-treatment or stepped care except PATS: see the trial table for details.
Control Group death rate per 1000 patients per year.
High doses studies were defined as those using starting drugs and doses greater than or equal to chlorthalidone 50mg, hydrochlorothiazide 50mg, chlorothiazide 500mg,
bendroflumethiazide 5mg, methychlothiazide 5mg 501.
Hypertension (partial update)
Trial
Pharmacological interventions
Table 51: Description of individual placebo controlled trials of thiazide-type diuretics
Hypertension (partial update)
Pharmacological interventions
A graphical presentation of pooled summary findings is shown in Figure 16 for all cause
mortality, fatal or non-fatal myocardial infarction (MI) and fatal or non-fatal stroke. The high
dose thiazide trials are of historical interest and, although the findings are more varied, the
overall summary for each endpoint is consistent with the findings from the low-dose thiazide
trials. The low dose trials show statistically significant reductions in mortality of 9%, in
myocardial infarction of 22% and in stroke of 31%: a statistically consistent finding across the
range of underlying risk.
Figure 16:
Meta-analysis of placebo-controlled randomised controlled trials of high
and low dose thiazide diuretics
Patients receiving placebo withdrew from treatment at an average rate of 10.7% per year.
Overall, withdrawal from active therapy was lower (Incident Risk Difference per year −1.2%,
95%CI: −1.9% to −0.6%) although there was variation between studies (Q, p<0.001).
Individual studies varied from a 4% reduction in withdrawal per year to no difference. While
rates of overall withdrawal are the most objective estimate of tolerability, they can conceal
different problems: lack of efficacy, perceived side-effects, adverse events or disease
progression. As the body of evidence increases in favour of new treatments some patients
may be withdrawn from placebo-controlled trials because of symptoms or signs indicating the
need for active therapy.
10.1.1.2
Beta-blockers
Seven trials with 27,433 participants were identified that met the review inclusion criteria (see
Table 52). Trials started between 1977 and 1988; enrolled patients had a mean age of 57
years, 49% were male and the mean duration of follow-up was 4.3 years. It is unclear what
proportion of participants was from ethnic minorities.
187
Beta-blocker1
Dose, mg
Country
Follow-up,
yrs
Start
year
Age in years
Range
Baseline BP,
mmHg
Number
enrolled
Baseline
Risk2
Mean
140
Atenolol
100
UK
4.4
1978
69
60–79
196/99
884
34
DUTCHTIA 19
IPPPSH 7
Atenolol
50
Netherlands
2.7
1986
-
-
158/91
1,473
29
Oxprenolol
160–320
International
3.4
1977
52
40–64
173/108
6,357
11
Propranolol
240
UK
4.9
1977
52
35–64
161/98
13,057
6
Atenolol
50–100
UK
5.8
1982
70
65–74
185/91
3,315
24
Beta-blocker or Diuretic3
Sweden
2.1
1985
76
70–84
195/102
1,627
37
Atenolol
Sweden
2.3
1988
70
40+
161/89
720
75
Coope
MRC
402
MRC-O
15
STOP-H
TEST 197
156
50
All trials featured stepped care, with additional drugs added if necessary
Control Group death rate per 1000 patients per year
Atenolol (50) or Metoprolol (100) or Pindodol (5)
Hypertension (partial update)
Trial
Pharmacological interventions
Table 52: Description of individual placebo controlled trials of beta-blockers
188
Hypertension (partial update)
Pharmacological interventions
A graphical presentation of pooled summary findings is shown in Figure 17 for all cause
mortality, fatal or non-fatal myocardial infarction (MI) and fatal or non-fatal stroke. Overall,
patients on beta-blockers had a statistically significant reduction in risk of stroke of 19%, and
non-significant reductions in risk of death of 6% and of myocardial infarction of 8%.
Figure 17:
Meta-analysis of placebo-controlled randomised controlled trials of
beta-blockers
Patients receiving placebo withdrew from treatment at an average rate of 10.6% per year.
Withdrawal per year from active therapy and placebo was similar (Incident Risk Difference
per year −0.4%, 95%CI: −1.6% to 0.8%) although there was variation between studies (Q,
p<0.001). Individual studies varied from a 5% reduction in withdrawal per year to a 2%
increase.
10.1.1.3
ACE inhibitors (ACEi)
One trial, with 6,105 participants and a mean follow-up of 3.9 years was identified that met
the review inclusion criteria (Table 53). The PROGRESS trial randomised patients following
stroke to perindopril with the addition of a diuretic (indapamide) if necessary or placebo.
Seventy percent of participants were male and 61% were Caucasian; 58% of patients assigned
to the ACEi also received the diuretic.
Table 53: Description of individual placebo controlled trials of ACEi
Trial
PROGR
ESS 500
ACEi 1
Perindopr
il
Dos
e,
mg
Country
4
Internation
al
Follo
w-up,
yrs
Star
t
year
Age in years
Rang
e
Mea
n
Baselin
e BP,
mmHg
Numbe
r
enrolle
d
Baselin
e Risk2
3.9
199
5
26–
91
64
147/86
6,105
27
The PROGRESS trial allowed physicians to add a diuretic if they deemed it appropriate
Control Group death rate per 1000 patients per year
PROGRESS did not show an overall reduction in mortality (RR 0.96, 95%CI: 0.83 to 1.12),
but statistically significant reductions in coronary events (RR 0.76, 95%CI: 0.60 to 0.96) and
stroke (RR 0.73, 95%CI: 0.64 to 0.84).
Patients receiving placebo withdrew from treatment during the PROGRESS trial at an average
rate of 8% per year. Withdrawal per year from active therapy was similar (Incident Risk
Difference per year 0.6%, 95%CI: −0.2% to 1.3%).
The recent HOPE25,652 study randomised patients with two or more cardiovascular risk factors
to a fixed dose of ramipril or placebo. The trial was designed similarly to trials of secondary
cardiovascular prevention rather than treatment of hypertension; the trial population were not
hypertensive and the study is not included in this review.
189
Hypertension (partial update)
Pharmacological interventions
10.1.1.4
Angiotensin receptor blockers
One trial, with 4,964 patients and a mean follow up of 3.7 years, was identified that met the
review inclusion criteria (see Table 54). The SCOPE trial randomised elderly patients with
mild to moderate hypertension and without cardiovascular disease in the preceding 6 months
to candesartan or placebo; approximately one third were male and ethnicity was not reported.
Table 54: Description of individual placebo controlled trials of angiotensin receptor
blockers
Trial
SCOP
E 371
ARB1
Candesartan
Dose
, mg
Countr
y
Follow
-up,
yrs
Start
year
8–16
Age in years
Rang
e
70–
89
Mean
Baselin
e BP,
mmHg
Number
enrolled
Baselin
e Risk2
Europe 3.7
199
76
166/90
4,964
29
and N.
7
Americ
a
Physicians could add a diuretic and other antihypertensive agents to patients in treatment or control groups if
they deemed it appropriate.
Control Group death rate per 1000 patients per year.
SCOPE did not show an overall reduction in mortality (RR 0.97, 95%CI: 0.83 to 1.14) or
coronary events (RR 1.10, 95%CI: 0.79 to 1.55), but a borderline statistically significant
reduction in stroke (RR 0.77, 95%CI: 0.59 to 1.01), primarily due to reduced non-fatal stroke.
Patients receiving placebo withdrew from treatment during the SCOPE trial at an average rate
of 8% per year. Withdrawal per year from active therapy was similar (Incident Risk
Difference per year −0.6%, 95%CI: −1.4% to 0.2%).
Two further placebo-controlled trials were identified (IDNT362 and RENAAL97), but not
considered adequately relevant to inform this guideline as both enrolled diabetic patients with
mild renal impairment.
10.1.1.5
Calcium-channel blockers
One trial, with 4,695 participants and median follow-up of two years, was identified that met
the review inclusion criteria (see Table 55). The SYST-EUR trial enrolled patients with
isolated systolic hypertension, one third of whom were male; ethnicity was not reported.
190
Hypertension (partial update)
Pharmacological interventions
Table 55: Description of individual placebo controlled trials of calcium-channel
blaockers
Trial
CCB1
Dos
e,
mg
Count
ry
Follo
wup,
yrs
Sta
rt
yea
r
Age in
years
Ran
ge
Mea
n
SYSTNitrendip 10– Europ 23
198 60+ 70
EUR43,124,207,555,558
ine
40
e
9
SYST-EUR featured stepped care, with additional drugs added if necessary.
Control Group death rate per 1000 patients per year.
Median follow-up.
Baseli
ne
BP,
mmH
g
174/8
6
Numb
er
enroll
ed
Baseli
ne
Risk2
4,695
27
SYST-EUR demonstrated no overall reduction in mortality (RR 1.06, 95%CI: 0.84 to 1.35),
some indication of a possible reduction in coronary events (RR 0.71, 95%CI: 0.45 to 1.10)
and a statistically significant reduction in stroke (RR 0.59, 95%CI: 0.41 to 0.84).
Patients receiving placebo withdrew from treatment at an average rate of 14% per year.
Withdrawal from active therapy per year was greater (Incident Risk Difference per year 2.3%,
95%CI: 0.8% to 3.9%).
Two further placebo-controlled trials were excluded because of uncertainty about the validity
of randomisation: SYST CHINA16,17,373,624] and STONE [233.
10.1.1.6
Alpha blockers
No placebo-controlled trials of alpha blockers in this patient group were identified that met
the review criteria.
10.2 2006 rapid pharmacological update: head to head trials
10.2.1
Most studies reported comparisons involving two or more drug classes in each treatment arm
administered according to a stepped administration protocol. In such cases, an initial
antihypertensive drug would be administered, followed by either:
 an increase in the dosage of the first drug, and/or
 the addition of a second drug if blood pressure targets were not reached using the first drug
alone.
All results should therefore be interpreted as demonstrating the efficacy and tolerability of
each drug only when used as the initial step in a wider antihypertensive drug treatment
regimen.
Many studies permitted a third drug to be added in patients unresponsive to both primary and
secondary antihypertensive drugs. Such drugs typically included alpha-blocking drugs such as
doxazosin or centrally acting antihypertensive drugs such as clonidine.
The update search found no new studies comparing ACEi or angiotensin-II receptor
antagonists with beta-blockers, or comparing ACEi with ARBs.
Three studies (CONVINCE78,79, NORDIL257,594 and CAPPP256,259,592) included in the original
guideline were excluded due to the confounded use of either beta-blocker or thiazide diuretic
as first-line antihypertensive therapy within the same treatment arm. A fourth study
(MAPHY)640 was a post-hoc follow-up of a subgroup of patients already included in the
HAPPHY study641, and so was excluded from the update.
One new study (MOSES)528 identified by the update search was excluded as it reported the
primary end-point as a composite of all-cause mortality, cardiovascular, and cerebrovascular
events, including all recurrent events, rather than as the first event only.
Clinical evidence statements: head-to-head drug comparisons
ACE inhibitors versus calcium-channel blockers
A meta-analysis of three studies (ALLHAT 589-591, JMIC-B650,651, STOP-H2155,255,258,368) comparing
ACE inhibitors with calcium-channel blockers (CCBs) showed that ACE inhibitors were associated
191
I
Hypertension (partial update)
Pharmacological interventions
with a higher incidence of stroke (RR 1.14, 95% CI 1.02 to 1.28) but a lower incidence of new-onset
diabetes (RR 0.85, 95% CI 0.75 to 0.98) and heart failure (RR 0.85, 95% CI 0.78 to 0.93). No
significant difference was found for mortality.
For MI there was substantial heterogeneity among the studies (I2 = 69%). Two studies (ALLHAT 589591
, JMIC-B650,651) found no significant difference between study drugs in terms of MI incidence,
while a third study (STOP-H2155,255,258,368) found that ACE inhibitors were associated with a reduced
incidence of MI (RR 0.77, 95% CI 0.62 to 0.96).
II
Of the two studies (ALLHAT 589-591, JMIC-B650,651) reporting the outcomes of unstable angina and
revascularisation procedures, neither found any significant difference.
The two studies (ALLHAT589-591, STOP-H2155,255,258,368) that reported the frequency of study drug
withdrawals each found ACE inhibitors to be associated with more withdrawals than CCBs
(respectively: RR 1.17, 95% CI 1.12 to 1.23; RR 1.14, 95% CI 1.06 to 1.24).
ARBs versus calcium-channel blockers
One study (VALUE)312 was found comparing ARBs with CCBs when used as first-line
antihypertensive therapy. ARBs were associated with a higher incidence of MI compared to CCBs
(RR 1.17, 95% CI 1.01 to 1.36). There was no significant difference in stroke reduction, mortality or
incidence of heart failure.
II
The study also reported frequencies of adverse events for each drug class and showed several
differences, but overall these did not particularly favour either drug. Pre-specified adverse events for
ARBs versus CCBs included peripheral oedema (14.9% versus 32.9%, p<0.0001), dizziness (16.5%
versus 14.3%, p<0.0001) and headache (14.7% versus 12.5%, p<0.0001). Additional adverse events
identified included diarrhoea (8.8% versus 6.8%, p<0.0001), serious cases of angina (4.4% versus
3.1%, p<0.0001) and syncope (1.7% versus 1.0 %, p<0.0001).
ACE inhibitors versus thiazide-type diuretics
A meta-analysis of three studies (ANBP2644, ALLHAT589-591, PHYLLIS657) comparing ACE
inhibitors with thiazide-type diuretics showed that ACE inhibitors are associated with a higher
incidence of stroke than thiazide-type diuretics (RR 1.13, 95% CI 1.02 to 1.25).
I
However, no difference was found for mortality.
For MI, the studies are heterogeneous (I2 = 66.5%). One study based in a relatively elderly and
predominantly white population (ANBP2)644 reported a lower incidence of MI for ACE inhibitors
(RR 0.71, 95% CI 0.51 to 0.98), but the remaining studies (ALLHAT 589-591, PHYLLIS657) found no
significant difference.
II
For heart failure, a meta-analysis of two studies (ALLHAT 589-591, ANBP2644) also demonstrated
heterogeneity (I2 = 67.1%). ALLHAT 589-591 reported a higher incidence with ACE inhibitors than
thiazide-type diuretics (RR 1.19, 95% CI 1.08 to 1.31), but in ANBP2 644 there was no significant
difference.
One study (ALLHAT)589-591 reported no significant difference in unstable angina but a higher
incidence of revascularisation procedures (RR 1.10, 95% CI 1.00 to 1.21) with ACE inhibitors.
Both studies (ALLHAT589-591 and ANBP2644) found ACE inhibitors to be associated with a higher
incidence of withdrawal compared to thiazide-type diuretics (RR 1.12, 95% CI 1.08 to 1.17; RR 1.10,
95% CI 1.04 to 1.17).
One study (ALLHAT)589-591 reported new-onset diabetes as an outcome, and found that the incidence
of diabetes after four years of follow-up was significantly higher for thiazide-type diuretics compared
to ACE inhibitors (p<0.001).
Calcium-channel blockers versus thiazide-type diuretics
A meta-analysis of five studies (ALLHAT 589-591, INSIGHT105,106, MIDAS90, NICS-EH343,
VHAS514,658) comparing calcium-channel blockers with thiazide-type diuretics found no significant
differences for mortality, MI or stroke. There was a statistically significantly higher incidence of
192
I
Hypertension (partial update)
Pharmacological interventions
heart failure with CCBs (RR 1.38, 95% CI 1.25 to 1.53).
Conversely, based on the results of three studies (ALLHAT589-591, INSIGHT105,106, NICS-EH343),
CCBs are associated with a reduced incidence of new-onset diabetes (RR 0.78, 95% CI 0.64 to 0.96).
Only the ALLHAT589-591 study reported unstable angina as an outcome and found no significant
difference between the drug classes. For revascularisation procedures, neither ALLHAT 589-591 nor
MIDAS90 found a significant difference.
II
In terms of study drug withdrawal, one study (INSIGHT) 105,106 found thiazide-type diuretics to be
associated with more withdrawals than CCBs (RR 1.20, 95% CI 1.13 to 1.28), although the other
studies (ALLHAT589-591, MIDAS90, VHAS514,658) did not find a significant difference between the two
drug classes.
Outcomes in those with isolated systolic hypertension (ISH)
A meta-analysis of three randomised controlled trials (SHEP 483,536,537,606, SHEP-P,281,484,485 SYSTEUR43,122,555) compared active antihypertensive drug therapy using either thiazide-based diuretics or a
calcium-channel blocker with placebo in patients with isolated systolic hypertension.
Antihypertensive drug therapy was associated with a reduced incidence of stroke (OR 0.62, 95% CI
0.51 to 0.77) and myocardial infarction (OR 0.74, 95% CI 0.61 to 0.91), although there was no
statistically significant difference in mortality rate.
I
Based on the results of a subgroup analysis from one randomised controlled trial (INSIGHT) 105,106,
initial antihypertensive therapy with the CCB nifedipine was comparable to the thiazide-type diuretic
hydrochlorothiazide plus amiloride in terms of mortality.
II
Based on the results of another subgroup analysis of patients with ISH from a randomised-controlled
trial involving patients with hypertensive LVH (LIFE) 328, initial therapy with an ARB is associated
with a reduced incidence of stroke (RR 0.60, 95% CI 0.38 to 0.92) and a lower mortality rate (RR
0.54, 95% CI 0.34 to 0.87) compared to initial antihypertensive therapy with a beta-blocker. The two
drugs were comparable in terms of the incidence of myocardial infarction.
Beta-blockers versus thiazide-type diuretics
641
402
Level
15
Three studies (HAPPHY , MRC , MRC-0 ) were found comparing the efficacy of beta-blockers
and thiazide-type diuretics. One study (HAPPHY) included only male patients.
I
A meta-analysis of these three studies showed no significant difference between the two drug classes
in terms of mortality.
Heterogeneity in the study results (I2 >75%) suggested that a meta-analysis would be inappropriate
for the outcomes of myocardial infarction and stroke. Sensitivity analyses were performed for
variation between the studies in terms of age (by including/excluding MRC-015, in which the average
age of participants was 70) and gender (by including/excluding HAPPHY) 641, but these were unable
to account for the observed heterogeneity.
II
One study (MRC-0)15 found beta-blockers to be associated with a higher incidence of myocardial
infarction compared to thiazide-type diuretics (RR 1.63, 95% CI 1.15 to 2.32). No association was
found in the other two studies402,641, which considered younger patients.
One study (MRC)402 in a relatively young population (average age 52 years) found beta-blockers to
be associated with a higher incidence of stroke compared to thiazide-type diuretics (RR 2.31, 95% CI
1.33 to 4.00). However, no association was found in the other two studies 15,641.
In terms of the frequency of withdrawal of the study drug, two studies (MRC 402, MRC-015) found
beta-blockers to be associated with more withdrawals (RR 1.06, 95% CI 1.01 to 1.11; RR 1.29, 95%
CI 1.22 to 1.37) while the remaining study641 reported a non-significant result.
Angiotensin-II receptor antagonists versus beta-blockers
One study (LIFE)176,222,507,618,619 was found comparing the angiotensin-II receptor antagonist (ARB)
losartan with the beta-blocker atenolol as first-line antihypertensive therapy.
The study found no significant difference between the two treatments in terms of myocardial
193
I
Hypertension (partial update)
Pharmacological interventions
infarction, revascularisation procedures, heart failure or angina. However, the study did find ARBs to
be associated with a reduced incidence of stroke (RR 0.75, 95% CI 0.63 to 0.88), new-onset diabetes
(RR 0.75, 95% CI 0.64 to 0.88) and fewer study drug withdrawals (RR 0.86, 95% CI 0.82 to 0.91).
Although mortality was lower in the ARB treatment group, this result was not statistically significant.
Calcium-channel blockers versus beta-blockers
A meta-analysis of three studies (ASCOT 157, ELSA656, INVEST481) compared calcium-channel
blockers (CCBs) with beta-blockers. There was no statistically significant difference in mortality or
myocardial infarction. Based on the results of the two studies reporting stroke as an outcome
(ASCOT157, ELSA656), CCBs were associated with a reduced incidence of stroke (RR 0.77, 95% CI
0.67 to 0.88).
I
For heart failure, a meta-analysis of two studies (ASCOT157, INVEST481) showed substantial
heterogeneity (I2 = 67.4%), but neither study alone found a statistically significant difference
between CCBs and beta-blockers.
II
Based on the results of one study (ASCOT)157, CCBs are associated with a reduced incidence of newonset diabetes (RR 0.71, 95% CI 0.64 to 0.78).
ASCOT157 also found CCBs to be associated with a lower incidence of unstable angina (HR 0.68,
95% CI 0.51 to 0.92) and fewer revascularisation procedures (HR 0.86, 95% CI 0.77 to 0.96) than
BBs, but the INVEST481 study found the association between both classes of drugs to be nonsignificant for these outcomes.
Study withdrawal was reported in two studies. In ASCOT 157 there were fewer withdrawals associated
with CCBs (RR 0.64, 95% CI 0.52 to 0.77), but in INVEST 481 there was no significant difference.
10.2.2
Meta-analysis results summary
Table 56 summarises the results from the meta-analysis comparing different drug classes in
general antihypertensive populations. Included are comparisons and outcomes in which interstudy heterogeneity was considered too great to include the pooled effect size in the evidence
statements above and hence these should be treated with caution.
Table 56: Summary of effect sizes for each comparison included in the meta-analysis
Studies
Total n
Effect size RR [95% CI]
I2
(%)
01 Mortality
3
15,765
1.04 [0.91, 1.20]
44.1
02 Myocardial infarction
3
15,765
1.15 [0.82, 1.60]
76.8
03 Stroke
3
15,765
1.27 [0.73, 2.23]
77.6
01 Mortality
1
9,103
0.89 [0.78, 1.01]
N/A
02 Myocardial infarction
1
9,103
1.05 [0.86, 1.28]
N/A
03 Stroke
1
9,103
0.75 [0.63, 0.88]
N/A
04 Heart failure
1
9,103
0.95 [0.76, 1.18]
N/A
05 Diabetes
1
7,998
0.75 [0.64, 0.88]
N/A
Comparison
01 Beta-blockers versus thiazides
03 ARBs versus beta-blockers
06 Calcium-channel blockers versus beta-blockers
01 Mortality
3
44,075
0.94 [0.88, 1.00]
5.7
02 Myocardial infarction (inc. silent MI)
3
44,075
0.93 [0.83, 1.03]
0
03 Myocardial infarction (exc. silent MI)
3
44,075
0.91 [0.81, 1.02]
0
04 Stroke
2
21,499
0.77 [0.67, 0.88]
0
05 Heart failure
2
41,833
0.96 [0.74, 1.26]
67.4
06 Diabetes
1
14,112
0.71 [0.64, 0.78]
N/A
194
Hypertension (partial update)
Pharmacological interventions
Comparison
Studies
Total n
Effect size RR [95% CI]
I2
(%)
04 ACE inhibitors versus calcium-channel blockers
01 Mortality
3
23,625
1.04 [0.98, 1.11]
0
02 Myocardial infarction
3
23,619
0.94 [0.74, 1.19]
69.3
03 Stroke
3
23,619
1.15 [1.03, 1.27]
5.2
04 Heart failure
3
23,619
0.85 [0.78, 0.93]
0
05 Diabetes
2
15,501
0.85 [0.76, 0.94]
15.2
01 Mortality
1
15,313
1.02 [0.93, 1.12]
N/A
02 Myocardial infarction
1
15,313
1.17 [1.01, 1.36]
N/A
02 Stroke
1
15,313
1.14 [0.97, 1.33]
N/A
03 Heart failure
1
15,313
0.88 [0.76, 1.01]
N/A
01 Mortality
2
29,697
1.00 [0.94, 1.06]
0%
02 Myocardial infarction
3
30,204
0.87 [0.60, 1.24]
66.5
03 Stroke
3
30,204
1.13 [1.02, 1.25]
0
04 Heart failure
2
29,697
1.07 [0.81, 1.41]
67.1
01 Mortality
5
32,195
0.97 [0.93, 1.02]
0
02 Myocardial infarction
5
32,195
1.02 [0.96, 1.08]
0
03 Stroke
5
32,195
0.93 [0.84, 1.04]
0
04 Heart failure
5
32,195
1.38 [1.25, 1.53]
0.2
05 Diabetes
3
20,885
0.82 [0.75, 0.90]
43.8
02 ARBs versus calcium-channel blockers
05 ACE inhibitors versus thiazides
07 Calcium-channel blockers versus thiazides
08 Antihypertensive therapy versus placebo (ISH population)
01 Mortality
3
9,745
0.88 [0.77, 1.01]
0
02 Myocardial infarction
3
9,745
0.75 [0.62, 0.91]
0
03 Stroke
3
9,745
0.64 [0.52, 0.78]
0
195
Hypertension (partial update)
Pharmacological interventions
10.3 2011 update: Pharmacological therapy for hypertension
10.3.1.1
Clinical evidence
The literature was reviewed from December 2005 onwards (this was the cut-off date of the
previous NICE guidance on pharmacological treatment of hypertension, CG34) for systematic
reviews and RCTs comparing ACEi vs ARB for first-line treatment in adults with primary
hypertension. RCTs were included if there was: ≥12 months follow-up, N≥200 and the
population did not consist of people who were exclusively diabetic or had CKD.
Three RCTs552,587,653 were found which fulfilled the inclusion criteria and addressed the
question and were included in the review.
196
Update 2011
10.3.1
Following the rapid pharmacological update of the guideline in 2006 the use of an algorithmbased approach to treatment was recommended, based on an A,C,D, where A represented an
ACEi (or ARB when an ACEi was not tolerated), C respresented a CCB, and D represented a
thiazide-type diuretic. The guideline also recommended that initial therapy for primary
hypertension (step 1) should be stratified according to age and ethnicity. Specifically, the
guideline recommended that for older people aged ≥55years, treatment should be initiated
with a CCB (C) or thiazide-type diuretic (D). For people under the age of 55 years, an ACEi
(or ARB id ACEi was not tolerated)(A) was recommended for initial (step 1) therapy. In the
absence of clinical outcomes data in younger people, this recommendation was based on data
suggesting that an ACEi (or ARB) was likely to produce the most effective blood pressure
lowering as initial therapy in younger patients. However, due a lack of head-to-head
comparison trials, it was unclear in 2006 whether an ARB could be considered equivalent to
an ACEi as intial therapy for younger people. The evidence review in 2006 had also
suggested that for black people of African and Caribbean descent at any age, a CCB or
thiazide type diuretic was the preferred initial therapy at any age.
Since 2006, important new data has become available in a number of areas; i) comparison of
ACEi with ARB – to determine if treatment with an ARB is equivalent at preventing clinical
outcomes when compared to treatment with an ACEi; ii) for step 2 therapy, comparison
between a a combination of A+C versus A+D on clinical outcomes – this is important because
if one of these combinations is preferred then it would impact on the preferred step 1 therapy
for people aged ≥55 years, or black people of African and Caribbean descent at any age; iii)
new data showing differential effects of antihypertensive treatments on blood pressure
variability, suggesting that blood pressure variability per se is an independent predictor of
clinical outcomes; iv) a review of diuretic therapy, specifically addressing whether the
predominant use of low dose bendroflumethiazide as the preferred diuretic for the treatment
of hypertension in the UK is justified when the majority of clinical trials have used different
thiazide-type diuretics; and v) new data on antihypertensive therapy options for resistant
hypertension (step 4 treatment). Finally, since 2006, the cost of antihypertensive therapies has
decreased significantly, some more than others (e.g. CCBs and ARBs) due to generics
becoming available. Consequently, this update of hypertension guideline dealing with
pharmacological treatment for primary hypertension reviewed recommendations with regard
to; i) the equivalence of ACEi versus ARBs on clinical outcomes; ii) the appropriate choice of
diuretic therapy for the treatment of hypertension and their place in the hierarchy of treatment;
iii) the preferred combination of therapies for step 2 and step 3 treatment; and iv) the
treatment of resistant hypertension, i.e. step 4 treatment. This review of pharmacological
treatment strategies was supported by an updated cost-effectiveness analysis comparing
different treatments with updated costings.
Angiotensin-converting enzyme inhibitors (ACEi) versus Angiotensin Receptor
Blockers (ARB)
Forest plots found in Appendix H: Forest plots.
Hypertension (partial update)
Pharmacological interventions
 The first RCT653 (the ONTARGET trial) compared treatment with the ACEi ramipril (5
mg/day) vs. the ARB telmisartan (50 mg/day) and vs. a combination of the two
(ACEi+ARB) in N=25,620 people with hypertension, and had a median follow-up time of
56 months. Treatment followed a stepped add-on therapy protocol (stepped up to double or
triple therapy) for non-responders in each arm.
 The second RCT587 compared treatment with the ACEi enalapril (20 mg/day) vs. the ARB
losartan (50 mg/day) in N=560 people with hypertension, and had a follow-up time of 24
months. Treatment followed a one-step dose adjustment protocol for the ACEi arm.
 The third RCT552 (CORD IB trial) compared treatment with the ACEi ramipril (5 mg/day)
vs. the ARB losartan (50 mg/day) in N=3860 people with hypertension, and had a followup time of 12 months. Treatment followed a stepped dose adjustment and add-on therapy
protocol (increased dose then if needed added on additional antihypertensive) for nonresponders in each arm.
NOTE: no quality of life data was found, or data assessing the effects of ACEi vs ARB in
people aged 80+ or black people of African and Caribbean descent.
NOTE: we additionally looked for outcomes relating to sexual dysfuntion in men, for ACE vs
ARB (as this is thought to be an important ussue particulary for erectile dysfunction
sufferers). However,no outcomes relating to this were reported in any of the studies.
10.3.1.2
Evidence statements - clinical
197
Update 2011
The evidence profile below (Table 57) summarises the quality of the evidence and outcome
data from the three RCTs552,587,653 included in this review, comparing ACEi versus ARB.
ARB was significantly better than ACEi for:
 less study drug withdrawals*
[moderate quality evidence]
There was NS difference between ACEi and ARB for:
 mortality (all cause)
[high quality evidence]
 MI (fatal and non-fatal)
[moderate quality evidence]
 stroke (fatal and non-fatal)
[moderate quality evidence]
 angina requiring hospitalisation [moderate quality evidence]
 coronary revascularisation
[high quality evidence]
 new onset diabetes
[moderate quality evidence]
 heart failure
[moderate quality evidence]
*There was significant heterogeneity for this outcome when the data from the three trials were
pooled together. Heterogeneity could be explained by the fact that both low and high quality
trials had been pooled together (details of sensitivity analysis by methodological quality can
be found in the forest plot for this outcome). Low quality trials were defined as those which
had no blinding or allocation concealment. Data included in GRADE for this outcome was
therefore based on the high quality trial alone. However the overall quality rating given by
GRADE for this outcome was ‘moderate’ due to imprecision (reasons outlined in the
evidence profile).
No of patients
No of
studies
Design
Limitations
Inconsistency
Indirectness
Imprecision
Effect
Relative
Other
considerations
ARB
ACEi
Quality
Absolute
(95% CI)
Mortality (all cause) (follow-up 12 - median 56 months)
2
CORDIB
552
ONTAR
GET653
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
995/10443
(9.5%)
1018/10535
(9.7%)
HR 0.98
(0.9 to
1.07)
2 fewer per
1000 (from
9 fewer to
6 more)
HIGH
417/10535
(4%)
HR 1.07
(0.94 to
1.22)
3 more per
1000 (from
2 fewer to
8 more)
MODERATE
413/10535
(3.9%)
HR 0.92
(0.8 to
1.06)
3 fewer per
1000 (from
8 fewer to
2 more)
MODERATE
925/8576
(10.8%)
HR 1.04
(0.95 to
1.14)
4 more per
1000 (from
5 fewer to
14 more)
MODERATE
1269/8576
(14.8%)
HR 1.02
(0.95 to
1.1)
3 more per
1000 (from
7 fewer to
14 more)
HIGH
MI (fatal and non-fatal) (follow-up 12-56 months)
2
CORDIB
552
198
ONTAR
GET653
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
2
none
443/10443
(4.2%)
Stroke (fatal and non-fatal) (follow-up 12 - median 56 months)
2
CORDIB
552
ONTAR
GET653
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious2
none
378/10443
(3.6%)
Hospitalisation for angina (follow-up median 56 months)
1
ONTAR
GET653
randomised
trials
no serious
limitations3
no serious
inconsistency
no serious
indirectness
serious2
none
954/8542
(11.2%)
Coronary revascularisation (follow-up median 56 months)
1
ONTAR
GET653
randomised
trials
no serious
limitations3
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
New onset diabetes (follow-up 12-56 months)
1290/8542
(15.1%)
Hypertension (partial update)
Summary of findings
Quality assessment
Pharmacological interventions
Table 57: Evidence profile comparing ACEi versus ARBs
ONTAR
GET653
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious2
none
404/7195
(5.6%)
372/7386
(5%)
HR 1.12
(0.97 to
1.29)
6 more per
1000 (from
1 fewer to
14 more)
MODERATE
537/8542
(6.3%)
514/8576
(6%)
HR 1.05
(0.93 to
1.19)
3 more per
1000 (from
4 fewer to
11 more)
MODERATE
2067/10665
(19.4%)
HR 0.87
(0.81 to
0.92)7
23 fewer
per 1000
(from 14
fewer to 34
fewer)
LOW
Heart failure (follow-up median 56 months)
1
ONTAR
GET653
randomised
trials
no serious
limitations3
no serious
inconsistency
no serious
indirectness
serious2
none
Study drug withdrawal (follow-up 12 - median 56 months)
1
ONTAR
GET653
1
randomised
trials
serious
3,4
no serious
inconsistency5
no serious
indirectness3
serious
6
none
1812/10572
(17.1%)
1/2 studies (CORD IB): no blinding, no allocation concealment; but this trial was small compared to the other included one (ONTARGET) so overall weighted as no serious limitations.
95% confidence interval includes both 1) no effect and 2) appreciable benefit or appreciable harm
Random, double blind, allocation concealment, powered, ITT analysis. However unclear final dropouts (but treatment withdrawal was <30% for median 56 months follow-up) so acceptable.
4
Patients who entered the trial had already been 'filtered' at run-in to exclude those with poor compliance or who did not perform well.
5
3 studies originally included and pooled but there was significant heterogeneity (p<0.1 and I2 >50%). Low quality trials removed based on sensitivity analysis, and result reported here is from the high quality trial
data.
6
95% confidence interval crosses both 1) no effect and 2) appreciable benefit or harm and non-appreciable benefit or harm
7
p<0.0001; favours ARB
2
3
Hypertension (partial update)
552
Pharmacological interventions
2
CORDIB
199
Hypertension (partial update)
Pharmacological interventions
10.3.1.3
Economic evidence
Three studies were identified in the update search that included ACEi and ARB in the
comparators but all were excluded due to being judged to have serious methodological
limitations202,529,560.
In the absence of a published cost effectiveness analysis, current UK drugs costs were
presented to the GDG to inform decision making. It was noted that losartan has recently come
off patent and other ARBs are also due to come off patent over the next few years.
10.3.1.4
Evidence statements – Clinical
ARB was significantly better than ACEi for:
 less study drug withdrawals*
[low quality evidence]
10.3.1.5
10.3.2
10.3.2.1
Evidence statements – Health economics
 No relevant evidence of cost-effectiveness was available.
 In terms of drug acquisition costs alone, in December 2010 based on BNF 60 the lowest
cost ARB was £25.94 per year (losartan [100mg used for costing]) and the lowest cost
ACEi was £20.73 per year (ramipril [10mg used for costing]).
Diuretics
In adults with primary hypertension, which is the most clinically and cost effective thiazide
type diuretic (bendrofluazide/bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide) for first line treatment, and does this vary with age and ethnicity?
Clinical evidence
Thiazide-type diuretics versus placebo or other antihypertensive drug class
The literature was searched for all years (as this was not addressed in the previous
guidelines)425,436. SRs/MAs and RCTs were included that compared the following TDs
(bendrofluazide/bendroflumethiazide, chlorthalidone, indapamide, hydrochlorothiazide) with
either placebo or other classess of a-HT drugs for 1st-line therapy. Studies were excluded if
they had sample sizes of N<200, follow-up of <1 year or populations which were exclusively
diabetic or had chronic kidney disease. Pre-specified outcomes of interest were only clinical
outcomes (e.g. stroke, MI etc.) and not BP measurements.
NOTE: in the previous NICE hypertension guidelines 425,436 a lot of the evidence for diuretics
was on Chlorthiazide, which is no longer used in the UK and is why many of the studies have
not been included in this review.
200
Update 2011
There was a non-significant difference between ACEi and ARB for:
 mortality (all cause)
[high quality evidence]
 MI (fatal and non-fatal)
[moderate quality evidence]
 stroke (fatal and non-fatal)
[moderate quality evidence]
 angina requiring hospitalisation [moderate quality evidence]
 coronary revascularisation
[high quality evidence]
 new onset diabetes
[moderate quality evidence]
 heart failure
[moderate quality evidence]
*There was significant heterogeneity for this outcome when the data from the three trials were
pooled together. Heterogeneity could be explained by the fact that both low and high quality
trials had been pooled together (details of sensitivity analysis by methodological quality can
be found in the forest plot for this outcome). Low quality trials were defined as those which
had no blinding or allocation concealment. Data included in GRADE for this outcome was
therefore based on the high quality trial alone. However the overall quality rating given by
GRADE for this outcome was still ‘low’ for reasons outlined in the evidence profile.
Hypertension (partial update)
Pharmacological interventions
14 RCTs (21 papers) were identified which fulfilled the inclusion criteria and addressed the
question, and were included in the review {1995 6420 /id;Sareli, 2001 489 /id;1978 6415
/id;Beckett, 2008 387 /id;The ALLHAT Officers and Co-ordinators for the ALLHAT
Collaborative Research Group, 2000 6139 /id;Weir, 2003 2500 /id;The Antihypertensive and
Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT), 2002 752
/id;Wing, 2003 6558 /id;Borhani, 1996 6140 /id;1985 1144 /id;Zanchetti, 2004 80
/id;Zanchetti, 1998 785 /id;Rosei, 1997 786 /id;Perry, 2000 417 /id;SHEP Cooperative
Research Group, 1991 470 /id;SHEP Cooperative Research Group, 1988 471 /id;Kostis, 1997
654 /id;Vaccarino, 2001 545 /id;Perry, 1986 418 /id;Hulley, 1985 6137 /id;Perry, 1989 6142
/id;Malacco, 2003 16093 /id;Tresukosol, 2005 1971 /id}. NOTE: several of the studies were
published as multiple papers (SHEP: three papers;335,483,606 SHEP-P: three papers;281,484,485
VHAS: two papers;514,658 and ALLHAT: three papers589,591,628) reporting different outcomes,
so these studies have only been counted once, however results from all the papers are reported
and referenced here483.
The table below (Table 58) summarises the studies included in the review. {1995 6420
/id;Sareli, 2001 489 /id;1978 6415 /id;Beckett, 2008 387 /id;The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2000 6139 /id;Weir, 2003 2500
/id;The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial
(ALLHAT-LLT), 2002 752 /id;Wing, 2003 6558 /id;Borhani, 1996 6140 /id;1985 1144
/id;Zanchetti, 2004 80 /id;Zanchetti, 1998 785 /id;Rosei, 1997 786 /id;Perry, 2000 417
/id;SHEP Cooperative Research Group, 1991 470 /id;SHEP Cooperative Research Group,
1988 471 /id;Vaccarino, 2001 545 /id;Perry, 1986 418 /id;Hulley, 1985 6137 /id;Perry, 1989
6142 /id;Malacco, 2003 16093 /id;Tresukosol, 2005 1971 /id}. Table 59 summarises the
diuretics used in each trial and their doses.
Data was categorised into those diuretics that were classed as:
 thiazide diuretics (TDs): bendrofluazide / bendroflumethiazide (BDZ) and
hydrochlorothiazide (HCTZ)
 ‘thiazide-like’ diuretics (TDLs): chlorthalidone (CTD) and indapamide (IND)
Table 58: Summary of included studies
Study
N
Intervention
Comparison
Follow-up
Results
17,354
BDZ
(10mg/day)
Propanolol
(240mg/day)
or
placebo
Mean 4.9
years
NS difference in overall
mortality, CHD events or
cardiovascular events
between BDZ and
propanolol. BDZ better
than propanolol for
reduced cerebrovascular
events.
NS difference in overall
mortality or CHD events
between BDZ and
placebo. BDZ better than
placebo for reduced
cardiovascular, and
cerebro-vascular events
200
HCTZ
(25-50
mg/day)
CCB
(amlodipine)
(5-10 mg/day)
18 months
No difference between
HCTZ and CCB for
mortality
TDs – BDZ
MRC8
TDs – HCTZ
THAI
elderly{Tresukos
ol, 2005 1971
/id}
201
Hypertension (partial update)
Pharmacological interventions
Study
N
Intervention
Comparison
Follow-up
Results
90
883
HCTZ
(25 – 50
mg/day)
CCB (isradipine)
(2.5- 5mg/daily)
36 months
NS differences between
HCTZ and isradipine for
overall mortality, CHD
events, cardiovascular, and
cerebro-vascular events
Sareli et al. 2001
409
HCTZ
(12.5
mg/day)
CCB (nifedipine
SR)
(30 mg/day)
or
CCB (verapamil
hydrochloride
SR)
(240 mg/day)
or
ACEi (enalapril
maleate)
(10 mg/day)
13 months
in total but
2 months
for
monotherap
y data
NS differences between
groups
508
HCTZ
(25 mg qid)
pravastatin in
50% of
patients.
ACEi (fosinopril)
(25mg qid)
pravastatin in
50% of patients.
Mean 2.6
years
NS differences in CHD
events, cerebrovascular
events or cardiovascular
events
VA-NHLBI3
1012
CTD
(50 mg/day
initially)
Placebo
2 years
NS differences between
groups
SHEP335,483,536,537,
4736
CTD
(12.5-25
mg/day)
Placebo
4.5 years
CTD better than placebo
for reduced CHD events,
reduced stroke and
reduced cardiovascular
events. NS difference for
HF (fatal and non-fatal).
SHEP- P281,484,485
441
CTD
(25-50
mg/day)
Placebo
34 months
NS differences between
groups
VHAS514,658
1414
CTD
(25mg/day)
CCB (verapamil)
(240mg/day)
2 years
NS differences in overall
mortality, CHD events, or
cerebrovascular
SHELL384
1882
CTD
(12.5-25
mg/day)
CCB (lacidipine)
(4-6 mg/day)
Median 32
months
No difference between
CTD and CCB for
mortality, stroke, MI and
HF
ALLHAT
42,418
CTD
(12.525mg/day)
CCB
(amlodipine)
(2.5- 10mg/day)
or
ACEi (Iisinopril)
(10-40mg/day)
Mean 4.9
years
NS difference between
CTD and ACEi I for
overall mortality and CHD
events. CTD better for
cardiovascular and
cerebro-vascular events
NS difference between
CTD vs. CCB for all cause
mortality and CHD events,
cardiovascular events, and
cerebrovascular events
MIDAS
524
PHYLLIS657
TDLs – CTD
606
589,591,628
202
Hypertension (partial update)
Pharmacological interventions
Study
N
Intervention
Comparison
Follow-up
Results
6083
CTD
(GP’s choice
of dose)
ACEi (enalapril)
(GP’s choice of
dose)
Mean 4.1
years
CTD worse than enalapril
for CHD events. NS
difference for overall
mortality, cardiovascular
and cerebro-vascular
events
PATS20
5665
IND
(2.5 mg/day)
Placebo
Mean 2
years
IND better for reduced
stroke (fatal and nonfatal), total mortality, CV
deaths and coronary
deaths
HYVET63
3845
IND SR
(1.5 mg/day)
Placebo
Mean 2.1
years
IND better for reduced MI
(fatal and non-fatal), HF
(fatal and non-fatal) and
mortality. NS difference
between groups for stroke
ANBP2
644
TDLs – IND
Table 59: Diuretic and dosage used in trial
Diuretic used
Number of trials
Doses used
5
Sareli524
ANBP2644
PHYLLIS657
MIDAS90
THAI elderly{Tresukosol, 2005 1971 /id}
12.5mg/day
At GPs discretion
25mg qid
25-50mg/day
25-50 mg/day
1
MRC8
10mg/day
2
PATS20
HYVET63
2.5mg/day
1.5mg/day (SR)
6
ALLHAT591,628
SHEP335,483,536,537
SHELL384
VHAS514,658
SHEP-P484,485
VA-NHLBI3
12.5 – 25mg/day
12.5 – 25mg/day
12.5-25 mg/day
25mg/day
25-50mg/day
50-100mg/day
TDs
HCTZ
BDZ
TDLs
IND
CTD
The evidence profiles below (Table 60 to
203
Hypertension (partial update)
Pharmacological interventions
Table 67) summarise the evidence and outcome data from the 14 RCTs{1995 6420 /id;Sareli,
2001 489 /id;1978 6415 /id;Beckett, 2008 387 /id;The ALLHAT Officers and Co-ordinators
for the ALLHAT Collaborative Research Group, 2000 6139 /id;Weir, 2003 2500 /id;The
Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHATLLT), 2002 752 /id;Wing, 2003 6558 /id;Borhani, 1996 6140 /id;1985 1144 /id;Zanchetti,
2004 80 /id;Zanchetti, 1998 785 /id;Rosei, 1997 786 /id;Perry, 2000 417 /id;SHEP
Cooperative Research Group, 1991 470 /id;SHEP Cooperative Research Group, 1988 471
/id;Kostis, 1997 654 /id;Vaccarino, 2001 545 /id;Perry, 1986 418 /id;Hulley, 1985 6137
/id;Perry, 1989 6142 /id;Malacco, 2003 16093 /id;Tresukosol, 2005 1971 /id} included in this
review comparing diureticsvs. placebo or other a-HT drug classes. Data are presented for
each diuretic.
NOTE: cerebrovascular events in some trials was cited and was synonymous with stroke.
204
Table 60: Bendroflumethazide versus placebo
Design
Limitations
Inconsistency
Indirectness
1
MRC8
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious2
none
128/3519 (3.6%)
Effect
control
Relative
(95% CI)
Absolute
Quality
253/6941 (3.6%)
HR 1 (0.81 to 1.24)
0 fewer per
1000 (from 7
fewer to 9
more)
LOW
HR 1 (0.8 to 1.25)
0 fewer per
1000 (from 7
fewer to 8
more)
LOW
109/6941 (1.6%)
HR 0.44 (0.30 to 0.63)
9 fewer per
1000 (from 6
fewer to 11
fewer)
LOW
352/6941 (5.1%)
HR 0.78 (0.65 to 0.94)
11 fewer per
1000 (from 3
fewer to 17
fewer)
LOW
CHD event (follow-up mean 4.9 years)
1
MRC8
randomised
trials
serious
1
no serious
inconsistency
no serious
indirectness
serious
2
none
119/3519 (3.4%)
234/6941 (3.4%)
Stroke (follow-up mean 4.9 years)
1
MRC8
randomised
trials
serious
1
no serious
inconsistency
no serious
indirectness
serious
3
none
18/3519 (0.5%)
Cardiovascular event (follow-up mean 4.9 years)
2
1
MRC8
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious3
none
140/3519 (4%)
1
0
5
Allocation concealment unclear and attrition high
95% CI includes no effect and appreciable benefit or appreciable harm
3
95%CI does not include no effect but crosses both appreciable benefit or harm and non-appreciable benefit or harm
2
Table 61: Indapamide versus placebo
Quality assessment
No of
studies
2
PATS20
HYVET63
Design
randomised
trials
Limitations
no serious
limitations1
Inconsistency
no serious
inconsistency
No of patients
Other
Indapamide
Indirectness
Imprecision
control
considerations
versus placebo
Overall mortality (follow-up mean 2.05 years)
no serious
indirectness
393/4736
(8.3%)
serious2
none
342/4774 (7.2%)
8.90%
CHD event (follow-up mean 2.05 years)
Summary of findings
Effect
Relative
Absolute
(95% CI)
HR 0.85
(0.74 to
0.99)
12 fewer per 1000
(from 1 fewer to
21 fewer)
13 fewer per 1000
(from 1 fewer to
22 fewer)
Quality
MODERATE
Hypertension (partial update)
No of
studies
No of patients
Other
Bendroflumethiazide
Imprecision
considerations
versus placebo
Overall mortality (follow-up mean 4.9 years)
Pharmacological interventions
Summary of findings
Quality assessment
no serious
limitations1
serious3
no serious
indirectness
serious2
none
50/4774 (1%)
1.90%
HR 0.53
(0.36 to
0.77)
LOW
HR 0.72
(0.61 to
0.87)
17 fewer per 1000
(from 8 fewer to
23 fewer)
16 fewer per 1000
(from 7 fewer to
22 fewer)
MODERATE
HR 0.77
(0.64 to
0.93)
12 fewer per 1000
(from 4 fewer to
19 fewer)
11 fewer per 1000
(from 3 fewer to
17 fewer)
MODERATE
HR 1.09
(1.03 to
1.16)
30 more per 1000
(from 11 more to
52 more)
MODERATE
Stroke (follow-up mean 2.05 years)
2
PATS20
HYVET63
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
286/4736
(6%)
serious2
none
210/4774 (4.4%)
5.70%
Cardiovascular event (follow-up mean 2.05 years)
2
PATS20
HYVET63
randomised
trials
1
PATS20
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
259/4736
(5.5%)
serious2
none
203/4774 (4.3%)
4.70%
Quality of life - no limitations in daily activities (follow-up mean 2 years)
1
2
2
3
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious2
none
2125/2841
(74.8%)
2019/2824
(71.5%)
Both had allocation concealment; attrition was >20% in one trial and no data provided in the other trial
95%CI does not cross the line of no effect but crosses both appreciable benefit or harm and non-appreciable benefit or harm
Heterogeneity was 77%. This could be due to different populations. One trial recruited adults aged 80 years+ and the other trial recruited patients with a recent TIA or stroke.
Hypertension (partial update)
randomised
trials
8 fewer per 1000
(from 4 fewer to
11 fewer)
9 fewer per 1000
(from 4 fewer to
12 fewer)
Pharmacological interventions
2
PATS20
HYVET63
78/4736
(1.6%)
0
6
Table 62: Chlorthalidone versus placebo
Quality assessment
Limitations
Inconsistency
3
SHEP335,483,536,537
SHEP-P484,485
VA-NHLBI3
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious2
none
8/508 (1.6%)
Quality
HR 0.87
(0.73 to
1.04)
1 fewer per
1000 (from 3
fewer to 0 more)
LOW
HR 2.0
(0.86 to
4.67)
16 more per
1000 (from 2
fewer to 56
more)
16 more per
1000 (from 2
fewer to 57
more)
VERY LOW
165/2479
(6.7%)
HR 0.63
(0.49 to
0.80)
24 fewer per
1000 (from 13
fewer to 33
fewer)
24 fewer per
1000 (from 13
fewer to 34
fewer)
MODERATE
0/504
(0%)
HR 4.31
(0.27 to
68.84)
0 more per 1000
(from 0 fewer to
0 more)
MODERATE
5/504
(1%)
CHD events (follow-up mean 2 years)
3
SHEP335,483,536,537
SHEP-P484,485
VA-NHLBI3
randomised
trials
serious1
serious3
no serious
indirectness
serious4
none
16/508 (3.1%)
8/504
(1.6%)
Stroke
2
2
SHEP335,483,536,537
SHEP-P484,485
randomised
trials
serious5
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
114/2808 (4.1%)
0
7
Cardiovascular event (follow-up mean 2 years)
2
SHEP335,483,536,537
VA-NHLBI3
1
randomised
trials
serious1,6
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
No ITT analysis conducted on data in one study, attrition >20% in two studies
95%CI crosses both no effect and appreciable harm or benefit
3
Heterogeneity 59%
4
95%CI does not cross no effect but includes both appreciable benefit or harm and non-appreciable benefit or harm
5
Attrition >20%
6
ITT analysis not conducted in one study and attrition > 20% in the other study
2
2/508 (0.4%)
Hypertension (partial update)
Design
Pharmacological interventions
No of studies
Summary of findings
No of patients
Effect
Relative
Other
Chlorthalidone
Indirectness
Imprecision
control
Absolute
considerations
versus placebo
(95% CI)
Overall mortality (follow-up mean 2 years)
Table 63: Chlorthalidone versus calcium channel blocker.
Summary of findings
No of patients
Effect
Relative
Other
Chlorthalidone
Indirectness
Imprecision
control
Absolute
considerations
versus CCB
(95% CI)
Overall mortality (follow-up 2 to 4.9 years)
4 more per 1000
1406/10439
(from 4 fewer to
(13.5%)
HR 1.03
12 more)
no serious
no serious
2329/16483
none
(0.97 to
indirectness
imprecision
(14.1%)
2 more per 1000
1.10)
7.50%
(from 2 fewer to
7 more)
CHD events (follow-up 2 to 4.9 years)
1 more per 1000
1474/9497
(from 7 fewer to
(15.5%)
HR 0.94
11 more)
no serious
no serious
2460/15543
none
(0.88 to
indirectness
imprecision
(15.8%)
1 more per 1000
1.0)
8.90%
(from 4 fewer to
7 more)
Stroke (follow-up 2 to 4.9 years)
Quality assessment
Limitations
Inconsistency
3
ALLHAT591,628
SHELL384
VHAS514,658
randomised
trials
serious1
no serious
inconsistency
2
ALLHAT591,628
VHAS514,658
randomised
trials
serious1
no serious
inconsistency
3
ALLHAT591,628
SHELL384
VHAS514,658
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious2
none
Quality
MODERATE
MODERATE
419/10439
(4%)
HR 0.94
(0.83 to
1.06)
2 more per 1000
(from 2 fewer to
8 more)
LOW
3941/14836
(26.6%)
2432/8790
(27.7%)
HR 0.96
(0.91 to
1.01)
12 more per
1000 (from 0
more to 23
more)
MODERATE
19/940 (2%)
23/942
(2.4%)
HR 0.83
(0.46 to
1.62)
4 fewer per
1000 (from 13
fewer to 15
more)
VERY LOW
14/940 (1.5%)
12/942
(1.3%)
HR 1.17
(0.54 to
2.53)
2 more per 1000
(from 6 fewer to
19 more)
VERY LOW
717/16483 (4.3%)
Cardiovascular events (follow-up mean 4.9 years)
2
1
ALLHAT591,628
randomised
trials
serious
3
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
0
8
Heart failure (follow-up mean 32 months)
1
SHELL384
randomised
trials
serious
4
no serious
inconsistency
no serious
indirectness
very serious
2,5
none
MI (follow-up mean 32 months)
1
SHELL384
1
randomised
trials
serious4
no serious
inconsistency
no serious
indirectness
very serious2,5
Attrition was >20% in both trials. There was inadequate explanantion of allocation concealment in one trial
95%CI includes both no effect and appreciable benefit or harm
3
Attirtion >20%
4
Unclear allocation concealment and open blind
5
95%CI includes both no effect and both appreciable benefit and appreciable harm
2
none
Hypertension (partial update)
Design
Pharmacological interventions
No of studies
Table 64: Chlorthalidone versus ACEi Inhibitor
Summary of findings
No of patients
Effect
Relative
Other
Chlorthalidone
Indirectness
Imprecision
control
Absolute
considerations
versus ACEi
(95% CI)
Overall mortality (follow-up 4.1 to 4.9 years)
2 more per 1000
1509/11822
(from 6 fewer to
(12.8%)
HR 1.00
9 more)
no serious
no serious
2413/17873
none
(0.94 to
indirectness
imprecision
(13.5%)
2 more per 1000
1.07)
10.70%
(from 5 fewer to
8 more)
CHD events (follow-up 4.1 to 4.9 years)
40 more per
1563/11822
1000 (from 6
(13.2%)
more to 81
HR 0.97
more)
no serious
no serious
2533/17873
none
(0.91 to
indirectness
imprecision
(14.2%)
29 more per
1.03)
1000 (from 5
9.50%
more to 60
more)
Stroke (follow-up 4.1 to 4.9 years)
4 fewer per
112/3044
1000 (from 1
(3.7%)
fewer to 8
HR 0.88
fewer)
no serious
2
serious
none
107/3037 (3.5%)
(0.79 to
indirectness
5 fewer per
0.98)
1000 (from 1
4.40%
fewer to 9
fewer)
Cardiovascular events (follow-up 4.1 to 4.9 years)
11 fewer per
394/3044
1000 (from 5
(12.9%)
fewer to 17
HR 0.91
fewer)
no serious
no serious
none
429/3037 (14.1%)
(0.86 to
indirectness
imprecision
17 fewer per
0.96)
1000 (from 7
20.80%
fewer to 26
fewer)
Quality assessment
Inconsistency
2
ALLHAT591,628
ANBP2644
randomised
trials
serious1
no serious
inconsistency
2
ALLHAT591,628
ANBP2644
randomised
trials
serious1
no serious
inconsistency
2
ALLHAT591,628
ANBP2644
randomised
trials
serious1
no serious
inconsistency
2
ALLHAT591,628
ANBP2644
randomised
trials
serious1
no serious
inconsistency
Quality
MODERATE
MODERATE
LOW
LOW
Hypertension (partial update)
Limitations
Pharmacological interventions
2
Design
0
9
No of studies
Table 65: Hydrochlorthiazide versus calcium channel blockers
Quality assessment
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
Quality
Overall mortality (follow-up 2 to 36 months)
3
Sareli, MIDAS, THAI{Sareli,
2001 489 /id;Borhani, 1996
6140 /id;Tresukosol, 2005 1971
/id}
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
very
serious2
none
10/599
(1.7%)
10/833
(1.2%)
HR 1.18
(0.48 to
2.90)
2 more per 1000
(from 6 fewer to
22 more)
VERY
LOW
HR 0.77
(0.37 to
1.57)
12 more per
1000 (from 7
fewer to 51
more)
11 more per
1000 (from 6
fewer to 46
more)
VERY
LOW
HR 1.99
(0.5 to
7.97)
13 more per
1000 (from 7
fewer to 90
more)
14 more per
1000 (from 7
fewer to 92
more)
VERY
LOW
HR 1.8
(0.94 to
3.44)
27 more per
1000 (from 2
fewer to 81
more)
23 more per
1000 (from 2
fewer to 69
more)
LOW
CHD events (follow-up 2 to 36 months)
19/733
(2.6%)
2
Sareli, MIDAS90,524
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
very
serious2
none
13/499
(2.6%)
2.30%
Stroke (follow-up mean 36 months)
6/442
(1.4%)
2
1
MIDAS90
randomised
trials
serious3
no serious
inconsistency
no serious
indirectness
very
serious2
none
3/441
(0.7%)
1.40%
1
0
Cardiovascular events (follow-up 2 to 36 months)
26/733
(3.5%)
2
Sareli, MIDAS90,524
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious4
none
14/499
(2.8%)
3%
1
None of the trials provide adequate information on allocation concealment. One of the trials had attrition >20% and ITT analysis was not conducted on the data in the other trial
95%CI includes no effect and appreciable benefit and appreciable harm
3
Trial did not provide adequate information on allocation concealment and attrition > 20%
4
95% CI includes both no effect and appreciable benefit or appreciable harm
2
Hypertension (partial update)
Design
Summary of findings
Effect
Relative
Absolute
(95% CI)
Pharmacological interventions
No of studies
No of patients
HCTZ
versus
control
CCB
Table 66: Hydrochlorthiazide versus ACEi Inhibitor
Quality assessment
Limitations
Inconsistency
Indirectness
serious1
no serious
inconsistency
no serious
indirectness
Imprecision
Other
considerations
(95% CI)
Absolute
Quality
Overall mortality (follow-up mean 2 months)
1
Sareli
524
randomised
trials
very
serious2
none
1/58 (1.7%)
0/60
(0%)
HR 4.06 (0.08
to 204.37)
0 more per 1000 (from
0 fewer to 0 more)
VERY
LOW
3/253
(1.2%)
1/254
(0.4%)
HR 3.02 (0.31
to 29.07)
8 more per 1000 (from
3 fewer to 104 more)
VERY
LOW
HR 3.90 (0.08
to 196.36)
11 more per 1000
(from 4 fewer to 535
more)
12 more per 1000
(from 4 fewer to 541
more)
VERY
LOW
HR 3.90 (0.08
to 196.36)
11 more per 1000
(from 4 fewer to 535
more)
12 more per 1000
(from 4 fewer to 541
more)
VERY
LOW
CHD events (follow-up mean 2.6 years)
1
PHYLLIS657
randomised
trials
serious3
no serious
inconsistency
no serious
indirectness
very
serious2
none
Stroke (follow-up mean 2.6 years)
1
PHYLLIS657
randomised
trials
serious3
no serious
inconsistency
no serious
indirectness
very
serious2
none
0/253 (0%)
1/254
(0.4%)
Cardiovascular event (follow-up mean 2.6 years)
2
1
PHYLLIS657
1
1
1
2
3
randomised
trials
serious3
no serious
inconsistency
no serious
indirectness
No information on allocation concealment and attrition >20%
95%CI includes both no effect and appreciable benefit and appreciable harm
No information on allocation concealment and unclear on attrition
very
serious2
none
0/253 (0%)
1/254
(0.4%)
Hypertension (partial update)
Design
Summary of findings
Effect
Relative
Pharmacological interventions
No of
studies
No of patients
HCTZ
versus
control
ACEi
Table 67: Bendroflumethiazide versus Beta blocker
Quality assessment
Limitations
Inconsistency
Indirectness
1
MRC8
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
very
serious2
none
control
Absolute
120/3558
(3.4%)
HR 1.08
(0.84 to
1.39)
3 more per 1000
(from 5 fewer to
13 more)
VERY
LOW
119/3519 (3.4%)
103/3558
(2.9%)
HR 1.17
(0.9 to 1.52)
5 more per 1000
(from 3 fewer to
15 more)
LOW
18/3519 (0.5%)
42/3558
(1.2%)
HR 0.43
(0.25 to
0.75)
7 fewer per 1000
(from 3 fewer to 9
fewer)
LOW
146/3558
(4.1%)
HR 1.03
(0.82 to 1.3)
1 more per 1000
(from 7 fewer to
12 more)
VERY
LOW
128/3519 (3.6%)
Quality
CHD events (follow-up mean 4.9 years)
1
MRC8
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious4
none
Stroke (follow-up mean 4.9 years)
1
MRC8
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
serious3
none
Cardiovascular events (follow-up mean 4.9 years)
1
MRC8
2
1
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
very
serious2
Allocation concealment unclear and attrition > 20%
95%CI includes both no effect and appreciable benefit and appreciable harm
3
95%CI does not include no effect but does cross appreciable and non-appreciable benefit and harm
4
95%CI includes no effect and appreciable benefit or appreciable harm
2
none
140/3519 (4%)
Hypertension (partial update)
Design
Effect
Relative
(95% CI)
Pharmacological interventions
No of
studies
Summary of findings
No of patients
Other
Bendroflumethiazide
Imprecision
considerations
versus Beta blocker
Overall mortality (follow-up mean 4.9 years)
1
2
Hypertension (partial update)
Pharmacological interventions
Table 68: Summary of included studies
Study
N
Intervention
Control
Follow-up
Results
Bowlus
196493
29
CTD
(50mg/day)
HCTZ
(100 mg/day
6 weeks
treatment, 2
weeks washout
NS difference in BP
between groups.
Ernst,
2006198
30
CTD
(12.5mg/day)
force titrated to
25mg/day
HCTZ
(25mg/day)
force titrated
to 50mg/day
8 weeks
treatment, 4
weeks washout,
8 weeks
treatment
NS difference (office BP
and 24hr ABPM) between
groups.
TDL vs TD
213
Update 2011
Head to head comparisons
The literature was searched for all years (as this was not addressed in the previous
guidelines)425,436. SRs/MAs and RCTs were included that compared the fllowing TDs with
each other: bendrofluazide/bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide for 1st-line therapy. There was no restriction placed on sample size or
follow-up time. Populations which were exclusively diabetic or had chronic kidney disease
were excluded. Outcomes of interest were only BP measurements. All studies included in this
review measured BP in the office. However two studies94,199 used both office and ABPM or
just ABPM measurements.
A total of 15 RCTs were found that fulfilled the inclusion criteria. The different comparisons
are detailed in the table (Table 1) below.
 Six RCTs 94,194,339,493,494,551 Emeriau, 2001195 were found which compared Indapamide
(IND) vs. Hydrochlorothiazide (HCTZ).
 Two RCTs 39,76 were found which compared Indapamide (IND) vs.
bendrofluazide/bendroflumethiazide (BDZ).
 Two RCTs 266,503 were found which compared Indapamide (IND) vs. chlorthalidone
(CTD).
 Three RCTs93 198 216 were found which compared Chlorthalidone (CTD) vs.
hydrochlorothiazide (HCTZ).
 One RCT5 was found which compared Hydrochlorthiazide (HCTZ) vs.
bendroflumethiazide (BDZ).
NOTE: several studies194,195,503 assessed additional arms treating people with other classes of
a-HT drugs. These were not included because they did not answer this part of the question
(TDs vs. TDs) and were not included in the first part of the question (TDs vs. placebo / other
a-HT classes) because they did not meet inclusion criteria (ie. were N<200 and/or had <1 year
follow-up time).
NOTE: all RCTs were underpowered to detect a difference in BP. In order to detect a 5mm
difference, a sample size of N≥500 is needed.
NOTE: five studies were cross-over trials: Bowlus 1964, Ernst 2006, Elliott 1991, Hatt 1975,
Kreeft 198493,194,198,266,339
The table below (Table 1) summarises the studies included in this review and the
results5,39,76,93,94,194,195,198,216,266,339,493,494,503,551
Data was categorised into those diuretics that were classed as:
 thiazide diuretics (TDs): bendrofluazide / bendroflumethiazide (BDZ) and
hydrochlorothiazide (HCTZ)
 ‘thiazide-like’ diuretics (TDLs): chlorthalidone (CTD) and indapamide (IND)
Hypertension (partial update)
Pharmacological interventions
Study
N
Intervention
Control
Follow-up
Results
Finnerty,
1976216
54
CTD
(50mg/day plus
placebo)
HCTZ
(100mg/day)
2 weeks no
treatment,
followed by 4
weeks of
treatment in
either arm.
NS difference in BP
between groups.
Kreeft,
1984339
17
IND
(2.5mg/day)
HCTZ
(50mg/day)
2 months
placebo run-in,
12 weeks TD
drug, 2 months
placebo
washout, 12
weeks alternate
TD drug.
NS difference in BP
between groups.
Plante,
1988493
47
IND
(2.5mg/day)
HCTZ
(50 mg/day)
48 weeks
IND better for reduced BP
(no P value reported) and
was less likely to be
associated with
hypokalaemia.
Plante,
1983494
24
IND
(2.5mg/day)
HCTZ
(50 mg/day)
4-6 washout
placebo period,
followed by 12
weeks active
therapy.
IND better for reduction in
DBP in the recumbent
position
Spence,
2000551
39
IND
(2.5mg/day)
HCTZ
(25 mg/day)
6 months
NS difference in BP
between groups
Brandao,
201094
94
IND
(1.5 mg/day)
HCTZ
(25 mg/day)
12 weeks
Previously
untreated
patients.
Addition of
ACEi at 6 weeks
if target BP not
met.
NS difference in BP
(office or ABPM) between
groups
Emeriau,
2001195
524
IND (SR)
(1.5 mg/day)
HCTZ
(25 mg/day)
4 week washout
placebo period;
12 weeks
treatment
Similar reduction in BP
between groups
(equivalence test)
Amlodipine
(5 mg/day)
Elliot,
1991194
11
IND (2.5mg/day)
or
HCTZ (25
mg/day)
Placebo
(lactose)
28 days
NS difference in BP
between groups.
Alem,
200839
26
IND
(2.5mg/day)
BDZ
(2.5 mg/day)
28 days
Both IND and BDZ
reduced BP to a
significant degree.
Bing,
198176
20
IND
(2.5mg/day)
BDZ
(5 mg/day)
22 weeks
Equivalent fall in BP in
both groups
214
Hypertension (partial update)
Pharmacological interventions
Study
N
Intervention
Control
Follow-up
Results
Rakić,
2002503
80
IND
(2.5mg/day)
CTD
(25mg/day)
NIC
(60mg/day)
PPL
(120mg/day)
6 months
Significant decreases in
BP in all groups
Hatt,
1975266
36
IND
(5mg/day)
CTD
(100mg/day)
10 days washout,
followed by 90
day crossover
IND better % reduction in
DBP.
44
HCTZ
(12.5mg/day)
BDZ
(12.5mg/day)
12 months
NS difference in BP
between groups.
TDL vs TDL
TD vs TD
Anonymou
s, 19845
Table 69: Thiazide drug and dosages used in trials
TD name
Number of trials
CTD
5
Bowlus, 196493
Ernst, 2006198
Finnerty, 1976216
Hatt, 1975266
Rakić, 2002503
HTCZ
Indapamide
Doses used
50mg/day
12.5mg/day force titrated to
25mg/day
50mg/day plus placebo
100mg/day
25mg/day
11
Anonymous, 19845
Elliot, 1991194
Bowlus, 196493
Ernst, 2006198
Finnerty, 1976216
Kreeft, 1984339
Plante, 1988493
Plante, 1983494
Spence, 2000551
Brandao, 201094
Emeriau, 2001195
12.5mg/day
25 mg/day
100mg/day
25mg/day force titrated to
50mg/day
100mg/day
50mg/day
50mg/day
50mg/day
25 mg/day
25 mg/day
25 mg/day
11
NOTE: ALL (except one) OF
THESE TRIALS STATED THAT
THE PREPARATION WAS SR.
ALL JUST STATED
INDAPMIDE AND THE DOSE.
1.5 mg/day
1.5 mg/day (SR)
2.5mg/day
2.5mg/day
2.5mg/day
5mg/day
2.5mg/day
2.5mg/day
2.5mg/day
2.5mg/day
Brandao, 201094
Emeriau, 2001195
Alem, 200839
Bing, 198176
Elliot, 1991194
Hatt, 1975266
Kreeft, 1984339
Plante, 1988493
Plante, 1983494
Rakić, 2002503
215
Hypertension (partial update)
Pharmacological interventions
TD name
Number of trials
Spence, 2000551
Doses used
2.5mg/day
BDZ
3
Alem, 200839
Bing, 198176
Anonymous, 19845
2.5 mg/day
5 mg/day
12.5mg/day
216
Table 70 to Table 75 below summarise the quality of the evidence and outcome data from the studies included in the review
39,76,93,94,194,195,198,216,266,339,493,503,551
Figure 1: TDL vs TD (CTD vs HCTZ)
Design
193
randomised
trials
193
randomised
trials
1198
randomised
trials
1198
randomised
trials
1198
randomised
trials
1216
randomised
trials
2
No of
studies
1
7
randomised
trials
1
NS differnce between groups
2
High dropout rates; no ITT analysis
3
unclear allocation concealment
1216
No of patients
Quality
MODERATE
MODERATE
MODERATE
MODERATE
MODERATE
MODERATE
MODERATE
Hypertension (partial update)
Summary of findings
Effect
Relative
Other
Limitations
Inconsistency
Indirectness
Imprecision
Chlorthalidone HCTZ
Absolute
(95%
considerations
CI)
SBP seated (change from baseline) BOWLUS (follow-up 6 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 7 lower ( to
serious
none
29
29
inconsistency
indirectness
imprecision
lower)1
DBP seated (change from baseline) BOWLUS (follow-up 6 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 2.1 lower ( to
serious2
none
29
29
inconsistency
indirectness
imprecision
lower)1
SBP seated (change from baseline) ERNST (follow-up 8 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 6.3 higher (
serious3
none
30
30
inconsistency
indirectness
imprecision
to lower)1
DBP seated (change from baseline) ERNST (follow-up 8 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 1.2 lower ( to
serious3
none
30
30
inconsistency
indirectness
imprecision
lower)1
SBP: 24h ABPM (change from baseline) ERNST (follow-up 8 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 5 lower ( to
serious3
none
30
30
inconsistency
indirectness
imprecision
lower)1
SBP unknown method (change from baseline) FINNERTY (follow-up 4 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 4 higher ( to
serious3
none
26
28
inconsistency
indirectness
imprecision
lower)1
DBP unknown method (change from baseline) FINNERTY (follow-up 4 weeks; measured with: mmHg; Better indicated by lower values)
no serious
no serious
no serious
MD 1.3 higher (
serious3
none
26
28
inconsistency
indirectness
imprecision
to lower)1
Quality assessment
Pharmacological interventions
Table 70: Thiazide-like diuretics versus thiazide diuretics (chlorthalidone versus hydrochlorthiazide)
Table 71: Thiazide-like diuretics versus thiazide-like diuretics (indapimide versus chlorthalidone)
Quality assessment
1266
randomised
trials
Limitations
Inconsistency
very serious1
no serious
inconsistency
Indirectness
Imprecision
Other
considerations
Indapamide versus
Chlorthalidone
Quality
SBP supine (end of follow-up) HATT (Better indicated by lower values)
no serious
indirectness
very
serious2
none
38
38
-
MD 0 higher (10.14
lower to 10.14
higher)
VERY LOW
38
-
MD 4 lower (9.94
lower to 1.94
higher)
VERY LOW
MD 3.10 higher
(3.08 lower to 9.28
higher)4
MODERATE
MD 3.50 higher
(0.22 lower to 7.22
higher)4
MODERATE
DBP supine (end of follow-up) HATT (Better indicated by lower values)
1
266
randomised
trials
very serious
no serious
inconsistency
1
no serious
indirectness
very
serious3
none
38
SBP supine (end of follow-up) RAKIC (follow-up 6 months; measured with: mmHg; Better indicated by lower values)
1503
randomised
trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious3
none
20
20
-
DBP supine (end of follow-up) RAKIC (follow-up 6 months; measured with: mmHg; Better indicated by lower values)
1503
randomised
trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious3
none
20
20
-
1
Although the trial was single blinded, randomisation and allocation concealment was not described and there was no ITT analysis
95%CI includes no effect and both appreciable benefit and appreciable harm
3
95%CI include no effect and appreciable benefit or harm
4
NS difference between groups
2
2
1
8
Table 72: Thiazide-like diuretics vs Thiazide diuretics (Indapamide versus hydrochlorthiazide)
Summary of findings
Effect
Relative
Other
Indapamide
Inconsistency
Indirectness
Imprecision
control
Absolute
(95%
considerations
versus HCTZ
CI)
SBP supine (end of follow-up) (follow-up 28 days to 48 weeks; Better indicated by lower values)
MD 8.36 lower
no serious
no serious
very serious2
none
77
74
(10.92 to 5.8
indirectness
imprecision
lower)
DBP supine (end of follow-up) (follow-up 28 days to 48 weeks; Better indicated by lower values)
MD 4.2 lower
no serious
no serious
serious3
none
77
74
(5.48 to 2.92
indirectness
imprecision
lower)
SBP upright (end of follow-up) (follow-up 28 days to 48 weeks; Better indicated by lower values)
MD 8.74 lower
no serious
no serious
very serious4
none
54
55
(11.75 to 5.73
indirectness
imprecision
lower)
Quality assessment
No of studies
Design
Limitations
5194,339,493,494,551
randomised
trials
serious1
5194,339,493,494,551
randomised
trials
very serious1
4194,339,494,551
randomised
trials
no serious
limitations
No of patients
Quality
VERY LOW
VERY LOW
LOW
Hypertension (partial update)
Design
Summary of findings
Effect
Relative
control
Absolute
(95%
CI)
Pharmacological interventions
No of
studies
No of patients
DBP upright (end of follow-up) (follow-up 28 days to 48 weeks; Better indicated by lower values)
2195,551
randomised
trials
no serious
limitations
very serious5
no serious
indirectness
no serious
imprecision
none
54
-
MD 3.85 lower
(5.41 to 2.28
lower)
LOW
-
MD 3.95 lower
(7.03 to 0.87
lower)
MODERATE
-
MD 0.76 lower
(2.5 lower to 0.98
higher)
MODERATE
-
MD 12.55 lower
(17.11 to 7.99
lower)
HIGH
21
-
MD 2.07 lower
(7.2 lower to 3.06
higher)
MODERATE
33
-
MD 5.5 higher (0
to 0 higher)9
MODERATE
33
-
MD 5.9 higher (0
to 0 higher)9
MODERATE
-
MD 7.5 higher (0
to 0 higher)9
MODERATE
-
MD 2.0 higher (0
to 0 higher)9
MODERATE
55
SBP supine (change from baseline) (follow-up 3-6 months; measured with: mmHg; Better indicated by lower values)
serious6
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
196
192
DBP supine (change from baseline) (follow-up mean 3-6 months; measured with: mmHg; Better indicated by lower values)
2195,551
randomised
trials
serious6
1551
randomised
trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
196
192
SBP upright (change from baseline) (follow-up mean 6 months; Better indicated by lower values)
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
18
21
DBP upright (change from baseline) (follow-up mean 6 months; Better indicated by lower values)
1551
randomised
trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious7
none
18
SBP seated (change from baseline) (follow-up 12 weeks; Better indicated by lower values)
1
94
randomised
trials
serious
8
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
32
2
DBP seated (change from baseline) (follow-up 12 weeks; Better indicated by lower values)
194
randomised
trials
serious8
1
9
194
randomised
trials
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
32
SBP: 24h ABPM (change from baseline) (follow-up 12 weeks; Better indicated by lower values)
serious8
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
32
33
DBP: 24h ABPM (change from baseline) (follow-up 12 weeks; Better indicated by lower values)
randomised
no serious
no serious
no serious
8
1
serious
none
32
33
trials
inconsistency
indirectness
imprecision
1
There were inadequate methodological information in two of the three trials
2
Heterogeneity was 78%
3
Heterogeneity was 76%
4
Heterogeneity was 72%
5
Heterogeneity 68%
6
1/2 studies unclear for allocation concealment
7
95%CI includes no effect and appreciable harm or benefit
8
unclear allocation concealment
9
There was NS differnce between groups
94
Hypertension (partial update)
randomised
trials
Pharmacological interventions
4194,339,494,551
Table 73: Thiazode-like diuretic versus thiazide diuretic (Indapamide vs benroflumethiazide)
Summary of findings
Quality assessment
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
Indapamide versus
Bendrofluazide/Bendroflumethiazide
Absolute
10
-
MD 32 lower
(72.34 lower
to 8.34
higher)
VERY
LOW
10
-
MD 2 lower
(32.58 lower
to 28.58
higher)
LOW
10
-
MD 5 lower
(18.85 lower
to 8.85
higher)
VERY
LOW
-
MD 0 higher
(30.97 lower
to 30.97
higher)
LOW
10
-
MD 5.6
higher (8.35
lower to 19.55
higher)
VERY
LOW
10
-
MD 3.2
higher (1.85
lower to 8.25
higher)
VERY
LOW
control
Quality
SBP supine (end of follow-up) (follow-up mean 22 weeks; Better indicated by lower values)
1
76
randomised
trials
very serious
no serious
inconsistency
no serious
indirectness
serious
none
10
SBP upright (end of follow-up) (follow-up mean 22 weeks; Better indicated by lower values)
176
randomised
trials
very
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
10
DBP supine (end of follow-up) (follow-up mean 22 weeks; Better indicated by lower values)
176
randomised
trials
very
serious1
no serious
inconsistency
no serious
indirectness
very serious2
none
10
DBP Upright (end of follow-up) (follow-up mean 22 weeks; Better indicated by lower values)
1
76
randomised
trials
very
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
10
10
2
SBP (absolute change) (follow-up mean 22 weeks; Better indicated by lower values)
2
0
1
39
randomised
trials
very
serious1
no serious
inconsistency
no serious
indirectness
serious
3
none
13
DBP (absolute change) (follow-up mean 22 weeks; Better indicated by lower values)
139
1
randomised
trials
very
serious1
no serious
inconsistency
no serious
indirectness
Lacked most methodological information
95%CI includes no effect and appreciable benefit and appreciable harm
3
95%CI includes no effect and appreciable and non-appreciable harm or benefit
2
serious3
none
13
Hypertension (partial update)
Design
Effect
Relative
(95%
CI)
Pharmacological interventions
No of
studies
No of patients
Table 74: Thiazide diuretic vs thiazide diuretic (hydrochlorthiazide vs bendroflumethiazide)
Summary of findings
Quality assessment
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
HCTZ
BDZ
Relative
Absolute
Quality
(95% CI)
SBP supine (change from baseline) (follow-up 12 months; measured with: mmHg; Better indicated by lower values)
15
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
21
15
-
MD 1 lower
(0 to 0
higher)2
MODERATE
MD 3 higher
(0 to 0
higher)2
MODERATE
MD 1 higher
(0 to 0
higher)2
MODERATE
MD 4 higher
(0 to 0
higher)2
MODERATE
DBP supine (change from baseline) (follow-up 12 months; measured with: mmHg; Better indicated by lower values)
15
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
21
15
-
SBP upright (change from baseline) (follow-up 12 months; measured with: mmHg; Better indicated by lower values)
15
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
21
15
-
DBP upright (change from baseline) (follow-up 12 months; measured with: mmHg; Better indicated by lower values)
2
15
randomised
trials
serious1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
21
15
-
Hypertension (partial update)
No of
studies
Effect
Pharmacological interventions
No of patients
2
1
Hypertension (partial update)
Pharmacological interventions
10.3.2.2
Economic evidence
No relevant economic studies were included that compared different types of diuretic.
Economic studies were considered relevant to the question if they compared one diuretic with
another or examine the impact of cost and effectiveness differences between different
diuretics on the overall decision about which drug to treat people with. Economic studies that
included only one type of diuretic were not considered helpful to decision making and were
excluded.
In the absence of a published cost effectiveness analysis, current UK drugs costs were
presented to the GDG to help inform decision making.
10.3.2.3
Evidence statements - Clinical
Diuretics versus placebo or other anti-hypertensive drugs
Table 75: Results of studies / meta-analysis
Class of
diuretic
Diuretic
name
Outcome measure and statistical significance (arm favoured)
MI
CV
event
Stroke
Mortality
CHD
event
HF
ADL
Studies /
references
Diuretics versus placebo
BDZ
SS (BDZ)
SS (BDZ)
NS
NS
MRC
TDLs
CTD
SS (CTD)
SS (CTD)
NS
SS (CTD)
SHEP,
SHEP-P,
VA-NHLBI
IND
SS (IND)
SS (IND)
SS (IND)
SS (IND)
SS (IND)
HYVET,
PATS
Diuretics versus other anti-hypertensive classes
TDs
TDLs
BDZ vs
BB
NS
SS (BDZ)
NS
NS
MRC
HCTZ vs
ACEi
NS
NS
NS
NS
PHYLIIS,
Sareli
HCTZ vs
CCB
NS
NS
NS
NS
Sareli,
MIDAS,
THAI
elderly
CTD vs
ACEi
SS (CTD)
SS (CTD)
NS
SS (CTD)
ALLHAT,
ANBP2
NS
NS
NS
NS
CTD vs
CCB
NS
222
NS
ALLHAT,
SHELL,
VHAS
Update 2011
TDs
Hypertension (partial update)
Pharmacological interventions
NOTE: Some data were not provided in a usable format for inclusion in meta-analysis or were
unable to be pooled; data from each of these studies has been summarised individually in
Table 68 (and in the evidence profiles), along with pooled data where meta-analysis was
possible.5,93,94,198,216,503
NOTE: all data given are for between-group differences
223
Update 2011 Update 2011
Head to head comparisons
NOTE: The results of the meta-analyses comparing IND vs HCTZ for SBP and DBP (supine
and upright) should be interpreted with extreme caution due to the observed significant
heterogeneity. This appears to be attributed to one of the RCTs494 which reports an effect size
in the opposite direction to the other studies and because it has much smaller SDs than the
other trials, it has therefore been weighted more highly. If this trial is removed from the MA
then heterogeneity is reduced to more acceptable levels of 0% and the effect becomes NS.
Removing the two lower quality trials (Plante, 1988 and Kreeft, 1984)339,493 from the analysis
did not result in removing the observed heterogeneity. If a random effects model is applied to
the pooled estimate, then the effect size also becomes NS.
Diuretic
name
(interventi
on)
Diuretic
name
(comparis
on)
Outcome measure and statistical significance (arm favoured)
Change from baseline
Supine
SBP
End of follow-up
Upright
DBP
SBP
DBP
NS
224
Thiazide-like diuretic vs thiazide-like diuretic
IND
CTD
NS
NS
24h ABPM
Supine
SBP
SBP
DBP
SBP
NS
NS
NS
NS
NS
NS
DBP
Upright
DBP
SBP
SBP
SBP
SBP
93,198,216
NS
SS*
(IND)
NS
SS*
(IND)
NS
NS
NS
SS*
(IND)
NS
SS*
(IND)
NS
94,194,195,339,49
3,494,551
NS
NS
39,76
266,503
TD vs TD
HCTZ
BDZ
NS
NS
NS
NS
5
*significant heterogeneity. Hetereogenity is removed if the Plante 2003 trial494 is excluded from the analysis, and the overall effect becomes
NS. If a random effects model is applied to the pooled estimate, then the effect size also becomes NS.
NOTE: there were no studies found that compared:
•
CTD vs BDZ
•
IND vs BDZ
Update 2011
Thiazide-like diuretic vs Thiazide diuretic
CTD
HCTZ
NS
(unclear BP
method)
IND
HCTZ
SS
NS
SS
(IND)
(IND)
IND
BDZ
Seated
Absolute
change
unclear method
Studies /
references
Hypertension (partial update)
Pharmacological interventions
Table 76: Results of studies / meta-analysis
Hypertension (partial update)
Pharmacological interventions
10.3.2.4
Evidence statements – Health economic
Update 2011
 No evidence comparing the cost-effectiveness of different diuretics was identified.
 In terms of drug acquisition costs alone, in December 2010 based on BNF 60:
bendroflumethiazide (2.5mg) cost £11.86 per year; chlortalidone (50mgd) cost £19.81 per
year; indapamide (2.5mg non-proprietary) cost £16.03 per year.
10.4 Cost-effectiveness analysis
10.4.1.1
Economic question
The aim of the model was to estimate the cost effectiveness of the various blood pressurelowering drug classes for the management of hypertension in primary care.
10.4.1.2
Population and subgroups
The model considered patients with essential hypertension seen in primary care, excluding
those with pre-existing cardiovascular disease (CVD), heart failure (HF) or diabetes. It was
designed to be run separately for different cohorts, defined by age (55, 65, 75 and 85) and sex.
In addition, the model classified these cohorts by baseline CVD risk (0.5%–5% per year), by
heart failure risk (0–5% per year) and by diabetes risk (0–5% per year). A base case analysis
was performed for 65-year-old men and women with 2% CVD risk, 1% HF risk and 1.1%
diabetes risk, and a sensitivity analysis considered the effect of varying these risk levels.
The trial evidence that the model is based on included relatively few younger (under 55) or
black people of African and Caribbean descent, so the results may not be reliable for these
groups. However, we did conduct sensitivity analyses to explore how different assumptions
about treatment effects might impact on the cost-effectiveness results for younger (45) and
black people of African and Caribbean descent.
10.4.1.3
Interventions compared
The analysis assessed the costs and effects of the various classes of blood pressure-lowering
drugs alongside a 'do nothing' comparator. Inclusion of no treatment as an option is important
for economic evaluations as it allows us to identify low-risk groups for whom treatment is not
likely to be cost effective.
The interventions compared were thus:
 no intervention (NI)
 thiazide-type diuretics (D)
 calcium-channel blockers (C)
 beta-blockers (B)
 ACEi/angiotensin-II receptor antagonists (ARBs) (A).
At basecase, it was assumed that 80% of patients starting on ACEi would continue with these,
but that 20% would switch to ARBs due to an inability to tolerate ACEi (expert opinion).
d
Note that 25mg was considered the optimal dose but only 50mg tablets were listed in the BNF.
225
Update 2011
10.4.1
This model was developed as part of the 2006 pharmacological update (CG34) to balance
clinical outcomes and to test the cost effectiveness of different classes of initial
antihypertensive medications. As part of the 2011 update this analysis was rerun with updated
costs. The relative risks for ARBs were also updated based on new ACEi vs ARB data.
A summary of the analysis methods and results are provided below. Full methods and results
including an overiew of the overall impact of the update compared to the previous analysis is
available in ‘Appendix I: Cost-effectiveness analysis – pharmacological treatment’.
Methodological introduction
Hypertension (partial update)
Pharmacological interventions
ACEi/ARBs were combined as a strategy as they were considered to have equivalent
effectiveness. The costs and effects of the drugs were weighted to take account of this.
For simplicity only first-line drugs were considered. However, it should be noted that the
relative treatment effects from the meta-analysis include additional benefits from various
second and third line treatments offered in the trials.
10.4.1.4
Outcomes
The treatment effects were measured in terms of prevention of CVD events (non-fatal
unstable angina, MI, heart failure and stroke) and CVD-related deaths. The only adverse
effects modelled were onset of HF and diabetes, although we did examine the possible impact
of other adverse reactions to the drugs in sensitivity analyses.
It should also be noted that the model does not explicitly include cost impacts of withdrawals,
non-concordance or transfers between treatments. The impact of such changes on
effectiveness is implicitly included through the use of intention-to-treat trial data.
Health outcomes for the cost-effectiveness analysis are summarised in the form of quality
adjusted life-years (QALYs), where one QALY represents one year of healthy life.
10.4.1.5
Cost effectiveness
The results of cost-effectiveness analysis are usually presented as incremental costeffectiveness ratios (ICERs), which determine the additional cost of using one drug (X) per
additional QALY gained, compared with no intervention or another drug (Y):
𝐶𝑜𝑠𝑡 𝑜𝑓 𝑋 − 𝐶𝑜𝑠𝑡 𝑜𝑓 𝑌
𝐼𝐶𝐸𝑅𝑠 =
𝑄𝐴𝐿𝑌 𝑜𝑓 𝑋 − 𝑄𝐴𝐿𝑌 𝑜𝑓 𝑌
Where more than two interventions are being compared, the ICERs are calculated using the
following process.
 The drugs are ranked in terms of cost, from the cheapest to the most expensive (cheapest
indicated by LC (lowest cost) in the results table below).
 If a drug is more expensive and less effective than the previous one, then it is said to be
ruled out by 'simple dominated' and is excluded from further analysis (indicated by ‘-‘ in
the results table below).
 ICERs are then calculated for each drug compared with the next most expensive nondominated option. If the ICER for a drug is higher than that of the next most effective
strategy, then it is ruled out by 'extended dominance' (indicated by ‘-‘ in the results table
below).
 ICERs are recalculated excluding any drugs subject to extended dominance (these ICERs
are given in the results table below).
It is important to bear in mind that comparison between the crude cost-effectiveness ratios for
two drugs each compared with 'no intervention' can be highly misleading. To illustrate, the
incremental cost of starting antihypertensive therapy with the cheapest drug is relatively low,
while the incremental benefit is high, and thus the ICER is small. A more expensive but more
effective drug may also appear to have a relatively small cost-effectiveness ratio when
compared with 'no treatment'. However, the more expensive drug may have a larger ICER
when it is compared with the cheaper drug – the incremental cost of switching from the
cheaper drug to the more expensive one may be quite large in relation to the incremental
health gain. Nevertheless, the more expensive drug may still be a cost-effective alternative to
the cheaper drug if its ICER is less than the maximum amount that we are prepared to pay for
a QALY, which is considered to be around £20,000 to £30,000 for NICE decisions. In this
situation the most cost-effective option is the more expensive drug, despite its larger ICER.
However, if the ICER for the more expensive drug were to exceed the threshold of £20,000 to
30,000 per QALY, then it would not be cost effective and the cheaper option should be
preferred.
226
Hypertension (partial update)
Pharmacological interventions
10.4.2
10.4.2.1
Results of the health economic model
Base case results
The base case results are presented in Table 3 for 65-year-old men and women with an annual
CVD risk of 2%, HF risk of 1% and diabetes risk of 1.1%. This analysis suggests that
antihypertensive treatment is cost effective for this population and that the most cost-effective
initial drug in this group is calcium-channel blockers (C). The ICER of C compared with
thiazide-type diuretics (D) is £1,520 to £1,960 per QALY gained, which is below the level
usually considered to be affordable in the NHS (about £20,000 to £30,000 per QALY).
Table 10.77: Base case results (65-year-old, 2% risk, 1.1% diabetes risk, 1% HF risk)
Men
Effect (QALYs)
ICER (£/QALY)
D
£3,910
10.22
LC
A
£4,010
10.21
-
C
£4,030
10.28
£1,960
B
£4,550
9.89
-
NI
£4,690
9.57
-
Cost (£)
Effect (QALYs)
ICER (£/QALY)
D
£4,310
10.65
LC
C
£4,390
10.71
£1,520
A
£4,400
10.63
-
B
£5,050
10.29
-
NI
£5,230
9.96
-
Women
Beta-blockers (B) are ruled out by simple dominance, since D, A and C are estimated to be
cheaper and more effective. This can be seen in Figure 1, since B lies to the northwest of D, A
and C. The ACEi/ARB option (A) is also ruled out by extended dominance, since treating
some patients with D and the remainder with C would be cheaper and more effective than A;
in Figure 18, A lies to the northwest of a straight line joining points D and C. However, it
should be noted that the absolute differences between A, C and D are small.
227
Update 2011
Cost (£)
Hypertension (partial update)
Pharmacological interventions
Figure 18:
Base case results (65-year-old, 2% cardiovascular risk, 1.1% diabetes
risk, 1% HF risk)
Women
£5,400
£4,800
£5,200
£4,600
£5,000
£4,400
£4,800
£4,200
£4,600
£1,960
£4,000
£3,800
9.40
9.60
9.80
10.00
10.20
10.40
£1,520
£4,400
£4,200
9.80
10.00
10.20
10.40
10.60
10.80
Update 2011
Mean cost (2009 UK £ per person,
discounted)
Men
Mean effect (QALYs per person,
discounted)
Mean effect (QALYs per person,
discounted)
No Intervention
Calcium-channel Blockers
ACE Inhibitors/Angiotensin II receptor antagonists
Thiazide-type Diuretics
Beta-blockers
QALYs = quality-adjusted life years
10.4.3
228
Update 2011
The results of this analysis are set out in more detail, together with the sensitivity analyses, in
‘Appendix I: Cost-effectiveness analysis – pharmacological treatment (updated 2011)’.
Conclusions
This analysis found that treating hypertension is highly cost-effective. Treatment resulted in
improved health outcomes (higher QALYs) with all of the drug classes in the model and
actually resulted in overall cost savings compared to no treatment as the reduction in
cardiovascular events led to savings that offset the relatively low cost of antihypertensive
medication; although it should be noted that this is based on low cost generic drugs. In most
people CCBs were found to be the most cost-effective treatment option for initial treatment of
essential hypertension.
In terms of how the analysis has changed in 2011 since 2006, the most significant change in
the model inputs in the 2011 update was the reduction in drugs costs; in particular the cost of
CCBs, ACEs and ARBs. CCBs remained the most cost effective option, meaning no change
from 2006 in the interpretation of the base-case result in terms of overall cost effectiveness.
The ICER for CCBs did however reduce considerably (from £12,250 to £1,960) making
CCBs more cost effective than they were in 2006. CCBs are also no longer the most
expensive option, both B and NI being more expensive, meaning that CCBs are now cost
saving compared to NI; this was not the case in the 2006 guideline. Another key difference is
that the absolute difference between ACEs/ARBs, CCBs and TDs is now much smaller than it
was in 2006 with BBs even less cost effective. The results of the subgroup analysis remain
largely unchanged apart from that in both men and women, CCBs are cost effective a greater
percentage of the time compared with TDs in higher CVD risk and older age groups; however
this difference is not very large. Both old and new analyses show similar trends of cost
effectiveness but the new analysis has ACE/ARB cost effective in fewer scenarios than before
with the heart failure risk where this is the case moving to intermediate/high risk.
The considerations that were highlighted in the 2006 guideline are still relevant and are
described below.
The trials on which the cost-effectiveness calculations are based did not, in general, show
large differences in clinical outcomes between drug classes. Some of the outcomes have point
estimates of effect that are not statistically significant. In these situations the point estimate is
used as the best estimate of effect and so effects that are not statistically significant have a
Hypertension (partial update)
Pharmacological interventions
bearing on the relative cost effectiveness. Where the outcomes have a large effect on quality
of life or cost (for example, stroke or death) the effect on overall cost effectiveness may be
relatively important. The GDG considered the effect of this uncertainty about important
outcomes in reaching their conclusions. The relative cost effectiveness of the agents also
depends on the propensity of patients treated with them to develop new-onset diabetes or
heart failure. The GDG were aware that both of these adverse outcomes should be treated
with some caution in this context. It is not clear that an elevated blood glucose developing as
a consequence of drug treatment has the same long-term health impact as in other
circumstances, and the same applies to heart failure diagnoses, particularly since the
definition of this outcome in some studies would not satisfy currently accepted criteria.
The applicability of the model to people under the age of 55 is uncertain, since it is based on
trial data from mostly older people. However, sensitivity analysis showed that the drugs that
affect the renin-angiotensin system are likely to be the most cost-effective option in this group
if they are even slightly more effective in the young than is suggested from the overall trial
data.
These results are sensitive to the cost of CCBs. The more expensive brands are not likely to
be cost effective for use in the NHS. For example, the model estimates that for 65-year-olds at
2% annual CVD risk, 1.1% diabetes risk and 1% heart failure risk CCBs are only cost
effective if they cost less than £94 per patient per year.
Finally, it should be emphasised that there is still considerable uncertainty about the size of
some treatment effects, which translates into uncertainty about the relative cost-effectiveness
of the drugs. The evidence base is also difficult to interpret because of the complex nature of
some of the treatment protocols and also because of differences in some of the trial
populations.
10.5 Step two therapy
10.5.1.1
Clinical evidence
229
Update 2011
The literature was reviewed from December 2005 onwards for systematic reviews and RCTs
comparing A+C versus A+D for second-line treatment in adults with primary hypertension.
RCTs were included if there was: ≥12 months follow-up, N≥200 and the population did not
consist of people who were exclusively diabetic or had CKD.
One RCT296 was found that fulfilled the inclusion criteria and addressed the question, and was
included in the review.
 The RCT296 (the ACCOMPLISH trial) compared treatment with the ACEi benazepril (20
then 40mg/day) + the CCB amlodipine (5 mg/day) vs. the ACEi benazepril (20 then
40mg/day) + the diuretic hydrochlorothiazide (12.5 mg/day) in N=11,506 people with
hypertension, and had a follow-up time of 24 months. Treatment followed a doseadjustment protocol for non-responders in each arm.
NOTE: no quality of life data was found, or data assessing the effects of ACEi vs ARB in
people aged 80+ or black people of African and Caribbean descent.
The evidence profile below (Table 78) summarises the quality of the evidence and outcome
data from the one RCT296 included in this review, comparing ACEi + CCB vs. ACE + D.
Quality assessment
No of
studies
1
No of patients
Other
A+C
considerations
Mortality (all cause): ACCOMPLISH trial (follow-up mean 36 months)
Design
Limitations
Inconsistency
Indirectness
randomised trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
Imprecision
serious
2
none
236/5744
(4.1%)
A+D
Summary of findings
Effect
Relative
Absolute
(95% CI)
Quality
262/5762
(4.5%)
HR 0.90
(0.76 to
1.07)
4 fewer per
1000 (from
11 fewer to 3
more)
MODERATE
159/5762
(2.8%)
HR 0.78
(0.62 to
0.99)4
6 fewer per
1000 (from 0
fewer to 10
fewer)
MODERATE
133/5762
(2.3%)
HR 0.84
(0.65 to
1.08)
4 fewer per
1000 (from 8
fewer to 2
more)
MODERATE
59/5762
(1%)
HR 0.75
(0.5 to 1.1)
3 fewer per
1000 (from 5
fewer to 1
more)
MODERATE
386/5762
(6.7%)
HR 0.86
(0.74 to 1)
9 fewer per
1000 (from
17 fewer to 0
more)
MODERATE
MI (fatal and non-fatal): ACCOMPLISH trial (follow-up mean 36 months)
1
randomised trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious3
none
125/5744
(2.2%)
Stroke (fatal and non-fatal): ACCOMPLISH trial (follow-up mean 36 months)
1
randomised trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious2
none
112/5744
(1.9%)
230
Hospitalisation for unstable angina: ACCOMPLISH trial (follow-up mean 36 months)
1
randomised trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
2
none
44/5744
(0.8%)
Coronary revascularisation: ACCOMPLISH trial (follow-up mean 36 months)
1
randomised trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious2
none
334/5744
(5.8%)
Study drug withdrawal: ACCOMPLISH trial (follow-up mean 36 months)
18 fewer per
1000 (from 5
fewer to 31
fewer)
1
Random, double blind, allocation concealment, powered, ITT analysis. However no washout / run-in and <20% drop-outs (but Tx withdrawal was >30% for median 36 months follow-up).
2
95% confidence interval includes both 1) no effect and 2) appreciable benefit or appreciable harm
3
95% confidence interval includes both 1) appreciable benefit or harm and 2) non-appreciable benefit or harm
4
p=0.04; favours A+C
5
p=0.01; favours A+C
1
randomised trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
serious3
none
1684/5744
(29.3%)
1756/5762
(30.5%)
HR 0.93
(0.88 to
0.98)5
MODERATE
Hypertension (partial update)
Pharmacological interventions
Table 78: ACEi + CCB versus ACEi +Diuretic for second line therapy – quality assessment
Hypertension (partial update)
Pharmacological interventions
10.5.2.1
Economic evidence
One study was identified in the update search that included A+C and A+D as comparators but
was excluded due to being judged to have serious methodological limitations522.
10.5.2.2
Evidence statements - clinical
10.5.2.3
Evidence statements – health economic
 No relevant cost-effectiveness evidence was identified.
231
Update 2011
ACEi + CCB was significantly better than ACEi + D for:
 MI (fatal and non-fatal)
[moderate quality evidence]
 less study drug withdrawals
[moderate quality evidence]
There was NS difference between A+C and A+D for:
 mortality (all cause)
[moderate quality evidence]
 stroke (fatal and non-fatal)
[moderate quality evidence]
 hospitalisation for unstable angina
[moderate quality evidence]
 coronary revascularisation
[moderate quality evidence]
 new onset diabetes
[moderate quality evidence]
Hypertension (partial update)
Pharmacological interventions
10.6 Resistant hypertension
The GDG agreed to define the term ‘resistant hypertension’ in the guideline as someone
whose blood pressure is not controlled to <140/90mmHg, despite optimal or best tolerated
doses of third line treatment.
10.6.1.1
Clinical evidence
232
Update 2011
The literature was searched for all years (as this was not addressed in the previous
guidelines)(Newcastle Guideline Development and Research Unit;National Collaborating
Centre for Chronic Conditions) and all study types were included. Studies were included that
compared 4th-line antihypertensive drugs with placebo,head to head comparisons or gave
before-and after data, in people with resistant hypertension (defined as: people whose blood
pressure remains uncontrolled, despite taking optimal doses of 3 anti-hypertensive drugs).
Populations which were exclusively diabetic or had chronic kidney disease were excluded.
Six cohort studies 126,163,226,347,383,511 were found which fulfilled the inclusion criteria and
addressed the question, and were included in the review.
 The first cohort study 163 identifed and categorised people with resistant hypertension
receiving treatment with spironolactone (‘true resistant hypertension), from people with
controlled (‘white coat reisistant’ hypertension). For those with ‘true resistant
hypertension’ the study then compared data from before to after the introduction of
spironolactone. The study had a total of N=236 participants and had a median follow-up
time of 15 months. Treatment began with an initial dose of 25mg, and was titrated to 50100mg/d as required.
 The second cohort study 347 assessed N=133 participants with resistant hypertension and
measured their blood pressure before and after spironolactone 25-50mg/d, with a 3-month
and 6-month follow up period.
 The third cohort study 383 compared two groups of people with hypertension (total of N=69
participants). Group A were untreated hypertensives and Group B were drawn from a
hypertension clinic with treatment resistant hypertension. Group A was randomised to
receive either spironolactone 50 mg/d or bendroflumethiazide 2.5 mg/d in a crossover
design. All people in group B received 50mg/d of spironolactone. Group A received four
weeks treatment, four weeks washout, four weeks treatment, and group B had a mean
follow up time of 3.7 months.
 The fourth cohort study 226 assessed N=12 people with resistant hypertension before and
after receiving spironolactone (25mg/d and force-titrated to 50mg/d at 4 weeks), and had a
follow up time of eight weeks treatment. Other anti-hypertensive treatment was
discontinued, if necessary for a low blood pressure.
 The fifth cohort study 126 reviewed participants with uncontrolled hypertension in the
ASCOT-BPLA open-label RCT. All participants N=1411 received an anti-hypertensive
regimen based on either Atenolol or Amlodopine. The comparison was between those who
were prescribed additional spironolactone vs. those who were not prescribed
spironolactone. The median follow up time was 5.5 years.
 The sixth cohort study 511 compared Spironolactone with Doxazosin in N = 198 patients
with resistant hypertension. There was no mean follow-up time reported. Participants were
followed up until treatment was changed with the addition of a new drug/change in dosage
to control blood pressure or when blood pressure was controlled within a pre-specified
target.
Hypertension (partial update)
Pharmacological interventions
No evidence profile was generated as GRADE was not performed in this guideline on
observational studies. However GRADE automatically assigns a quality rating of ‘low’ to
observational studies.
The table below (Table 79) summarises the quality of the evidence and the outcome data from
the six cohort studies 126,163,226,347,383,511 included in this review of the effectiveness of 4th line
antihypertensive treatment in resistant hypertension in adults.
Table 79: Summary table of studies examining the role of fourth line antihypertensives
in resistant hypertension
Study
Intervention
Comparison
Follow-up
Results
Rodilla et
al.
2009{Rodil
la, 2009
16014 /id}
Spironolactone
Doxazosin
Until change
of treatment/
target blood
pressure
maintained
Spironolactone best
(decreased home or
ambulatory SBP and
DBP)
Low
Mahmud et
al.
2005{Mah
mud, 2005
15968 /id}
Previously untreatedspironolactone/bendrofl
umethiazide
4th line
Spironolacton
e
3-4 months
Spironolactone
effective in reducing
BP when used as a
4th line drug
Low
Chapman
et al.
2007{Chap
man, 2007
373 /id}
ASCOT trial patients
an a-HT regimen based
on either Atenolol or
Amlodopine
Plus
addition of
Spironolactone
ASCOT trial
patients on aHT regimen
based on
either
Atenolol or
Amlodopine
Median 5.5
years
Addition of
spironolactone
effective in reducing
BP
Low
De Souza
et al.
2010{de
Souza F.,
2010
15965 /id}
Spironolactone
Before vs.
after
Spironolacton
e
12 months
(Median 15
months, IQR
13-20 months)
Spironolactone
effective in reducing
‘office’ and
ambulatory blood
pressure.
Low
Lane et al.
2007{Lane,
2007 802
/id}
Spironolactone
Before vs.
after
Spironolacton
e
6 months
Spironolactone
effective in reducing
SBP and DBP
Low
Gaddam et
al.
2010{Gadd
am, 2010
15967 /id}
Spironolactone
Before vs.
after
Spironolacton
e
8 weeks
Addition of
spironolactone
effective in reducing
SBP and DBP
Low
Economic evidence
No relevant economic studies were identified that examined drugs in patients with resistant
hypertension.
In the absence of a published cost effectiveness analysis, current UK drugs costs for agents
that might be considered for use in resistant hypertension were presented to the GDG to help
inform decision making.
10.6.1.3
Evidence statements – clinical
Six studies found that blood pressure was reduced in people with resistant hypertension who
were treated with 4th-line spironolactone.
233
Update 2011
10.6.1.2
Evidenc
e
Quality
Hypertension (partial update)
Pharmacological interventions
10.6.1.4
Evidence statements – economic
 No relevant cost-effectiveness evidence was identified.
 In terms of drug acquisition costs alone, in December 2010 based on BNF 60:
spironolactone (25mg) cost £23.73 per year.
234
Update 2011Update 2011
One study 511 found that 4th line therapy with spironolactone was better than doxazosin for
reduction in SBP and DBP [low quality]
Three studies163,347 226 found that SBP and DBP was reduced after 4th line spironolactone
treatment (vs. before treatment). [low quality].
One study 383 found BP reduced in those treated with spironolactone compared with those
previously untreated and reported drop out rates of 10% due to adverse effects [low quality].
One study 126 found the addition of spironolactone (as 4th line therapy) was effective in
reducing BP, and an adverse event rate of 13% was reported [low quality].Evidence
statements – health economic
Hypertension (partial update)
Pharmacological interventions
10.7 Special groups for consideration
10.7.1
10.7.2
People aged over 80 years
See section 8 on page 106.
Younger people
Outcomes in younger patients
The literature search found no evidence for the clinical outcomes summarised above, therefore blood
pressure response to drug therapy was used as a surrogate. Three studies 164,177,394 and an age-stratified
analysis from a fourth study55 compared blood pressure response across various drug classes and
identified ACE inhibitors and beta-blockers as more effective at lowering blood pressure in younger
people, when compared to calcium channel-blockers or thiazide-type diuretics.
In older people, initial treatment with calcium channel-blockers or thiazide-type diuretics has been
shown to be more effective at blood pressure lowering than ACE inhibitors, angiotensin-II receptor
antagonists or beta-blockers157,312,589-591.
10.7.3
Ethnicity
There are ethnic differences in the prevalence of high blood pressure. In African American
patients, the prevalence of hypertension and mortality arising from complications such as
cardiovascular, cerebrovascular and renal disease is higher than other ethnic groups.
40,110,127,145,542
Mortality data from England and Wales (1988–92) shows similar trends, with
mortality due to hypertensive complications 3.5 times higher than the national average in the
African-Caribbean population. 504 British Asians also exhibit hypertension associated
mortality rates 1.5 times higher than the national average. 504
The Whitehall II Study investigated a cohort of London-based civil servants aged 35–56
years, between 1985 and 1988.638 A 73% response rate provided a cohort including 8,973
white participants, 577 of South Asian origin and 360 of African-Caribbean origin.
Participants were considered hypertensive if they had blood pressure above 160/95 mmHg or
were receiving antihypertensive drugs. African-Caribbean (odds ratio: 4.0; 95%CI: 2.8 to 5.7)
and South Asian (odds ratio: 2.3; 95%CI: 1.6 to 3.3) participants had a greater prevalence of
hypertension than white participants, after findings were adjusted for age, service grade, sex
and body mass index. Similarly, diabetes was more common in African-Caribbean
(unadjusted odds ratio: 2.8; 95%CI: 1.7 to 4.6) and South Asian (unadjusted odds ratio: 4.2;
95%CI: 3.0 to 5.8) participants. Although both ethnic groups had lower total cholesterol
scores that white participants, South Asian people tended to have a poorer lipid profile while
African-Caribbean people tended to have a more favourable one.
A study conducted in nine practices in South London interviewed men and women aged 40–
59 years of white, African and South Asian origin.116 Random samples of each group were
invited: 64% took some part in the study, although only about one half of these contributed
blood pressure data. As with the Whitehall study, individuals were considered hypertensive if
they had blood pressure above 160/95 mmHg or were receiving antihypertensive drugs. Age
and sex adjusted prevalence ratios for hypertension were 2.6 (95% CI: 2.1 to 3.2) in people of
African descent and 1.8 (95% CI: 1.4 to 2.3) in those of South Asian descent. Diabetes
prevalence ratios were 2.7 (95% CI: 1.4 to 2.3) and 3.8 (95% CI: 2.6 to 5.6) for those of
African and South Asian descent respectively. Differences in ethnic groups (West African vs.
Caribbean and Hindu vs. Muslim) were not statistically significant. Similarly to the Whitehall
study, people from these ethnic minority groups had lower total cholesterol scores than white
participants although a lipid profile was not attempted.
A number of other studies of local populations have explored the relationship between
ethnicity and cardiovascular risk factors. These studies raise methodological issues and do not
provide a useful picture of hypertension because they did not seek to adjust for treatment.
They demonstrate that varying patterns of risk factors may occur in different groups, although
235
Hypertension (partial update)
Pharmacological interventions
these may only be well understood with more definitive epidemiological research. A study
comparing South Asian and European participants in Newcastle upon Tyne found that
Bangladeshi participants had the poorest lipid profile while Indians had the best, similar to a
European profile.74,286 The age-adjusted prevalence of diabetes varied between Bangladeshi
(23%), Pakistani (23%), Indian (13%) and European (4%) participants. A London based study
drawing from factory worker and general practice populations confirmed the findings of the
Whitehall II study, showing similar trends in lipid profile comparing European, South Asian
and African-Caribbean participants. 400 Similarly a raised age-adjusted prevalence of diabetes
was seen in Sikh (20%), Punjabi Hindu (19%), Gujarati Hindu (20%) and Muslim (19%)
groups compared to white participants (5%). A survey of Bangladeshi participants in East
London found a poor lipid profile and raised prevalence of diabetes compared to a non-Asian
population.399
The evidence thus shows that hypertension and diabetes are more common among certain
ethnic groups in the UK. This greater prevalence of hypertension may lead to higher rates of
cardiovascular disease and target organ damage.145,230,236,252,409,542 Reasons for this greater
prevalence may be environmental as well as physiological. A trend towards increased blood
pressure and weight was observed with increasing urbanisation of rural black Africans496, and
with the migration of Punjabi participants from India to England.73
10.7.3.1
Clinical evidence
Update 2011
236
Update 2011
The literature was reviewed from December 2005 onwards (the cut-off date of the previous
guideline, CG34,425 where this was covered previously) for systematic reviews, RCTs, subgroup analyses of RCTs and cohort studies looking at first-line anti-hypertensive treatment of
black people of African or Caribbean descent who have primary hypertension. Studies were
included if there was: N≥1000 and the population did not consist of people who were
exclusively diabetic or had CKD.
Two subgroup analyses354,492 of an RCT (ALLHAT) were found which fulfilled the inclusion
criteria and addressed the question, and were included in the review. The ALLHAT study was
originally included in the previous NICE guidelines.425,441 ALLHAT compared ACEi vs TD
vs. CCB vs. alpha-blocker and 1/3 of the population were black people (NOTE: the term
‘black’ was that used in the ALLHAT trial). However, the studies included in the previous
guidelines did not give data for the ACEi vs. CCB arms in black people and did not give the
incidences of angioedema, which these newer subgroup analyses have looked at. Both the
subgroup analyses were planned a-priori as part of the design of the ALLHAT trial.
 The first subgroup analysis of the ALLHAT RCT492 assessed the incidence of angioedema
in people treated within each arm of trial (ACEi vs. TD vs. CCB vs. alpha-blocker) and the
incidence of the outcome in different subgroups of people (including different ethnic
groups: black people vs. non-black people). The study follow-up time was mean 4.9 years
and the number of people who developed angioedema was N=53 out of the total study
group of N=42,418. Because the data we are interested in is the incidence of agioedema in
black people vs. non-black people (ie. has come from the subgroup analysis), this study
data has been classed as ‘observational’ (see section below entitled ‘evidence profile’).
 The second sub-group analysis of the ALLHAT RCT354 assessed the incidence of clinical
endpoints that occurred in subgroups of patients, including black people vs. non-black
people who were randomised to the ACEi and CCB arms of the ALLHAT trial. The study
follow-up time was mean 4.9 years and the number of people who developed angioedema
was N=53 out of the total study group of N=42,418. This study has been classified as
‘observational’ because it is a subgroup analysis of an RCT.
The evidence profiles below (Figure 1 and Figure 2) summarises the quality of the evidence
and outcome data from the two RCT (ALLHAT) subgroup analyses354,492 included in this
Hypertension (partial update)
Pharmacological interventions
review, comparing outcomes in black people and non-black people. Where data was unable
to be put into GRADE, it has been written up narratively in the evidence statements.
237
Summary of findings
Quality assessment
No of patients
No of
studies
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
ACEi
Effect
other a-HT
classes (TD,
CCB or
alpha)
Relative
Quality
Absolute
(95% CI)
Angioedema (black people) out of total randomised (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
23/3210
(0.7%)
6/10196
(0.1%)
RR 12.18 (4.96 to
29.88)
7 more per 1000
(from 2 more to
17 more)
HIGH
1 fewer per
1000 (from 1
more to 1
fewer)
MODERAT
E
375 fewer per
1000 (from 375
fewer to 375
fewer)
MODERAT
E
625 fewer per
1000 (from 625
fewer to 625
fewer)
MODERAT
E
Angioedema (non-black people) out of total randomised (follow-up mean 4.9 years)
238
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
2
none
23/3210
(0.7%)
6/10196
(0.1%)
RR 0 (2.47 to 0)
3
Angioedema (black people) out of those who developed angioedema (follow-up mean 4.9 years)
1
randomised
trials
serious
4
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
23/37
(62.2%)
6/16 (37.5%)
inappropriate to
calculate (loss of
randomisation)
Angioedema (non-black people) out of those who developed angioedema (follow-up mean 4.9 years)
1
1
randomised
trials
serious
4
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
14/37
(37.8%)
10/16
(62.5%)
inappropriate to
calculate (loss of
randomisation)
Subgroup analysis of RCT: but pre-specified and the trial deliberately recruited a specific number of black people to be able to do this analysis
95% confidence interval excludes no effect, but the CI includes appreciable benefit and non-appreciable benefit or appreciable harm and non-appreciable harm
3
SS - favours other a-HT classes (p<0.0001)
4
Loss of randomisation in groups (incidence of angioedema in black people and non-black people, out of those who developed angioedema in the trial, rather than all participants randomised in the trial)
2
Hypertension (partial update)
Pharmacological interventions
Table 80: Evidence profile comparing ACEi versus other antihypertensive classes (TD, CCB or alpha) in black people and non-black
people (data from Piller et al., 2006)492
Summary of findings
Quality assessment
No of patients
No of
studies
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other
considerations
Effect
Relative
ACEi
CCB
Quality
Absolute
(95% CI)
CHD (black people) (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
no serious
imprecision
none
data not given in
study
1.09 (0.92,
1.03)
not enough
data given in
study to
calculate
HIGH
data not given in
study
0.97 (0.86,
1.10)
not enough
data given in
study to
calculate
MODERATE
data not given in
study
1.51 (1.22,
1.86) 5
not enough
data given in
study to
calculate
MODERATE
data not given in
study
1.07 (0.89,
1.28)
not enough
data given in
study to
calculate
LOW
data not given in
study
1.13 (1.02,
1.24)5
not enough
data given in
study to
calculate
MODERATE
CHD (non-black people) (follow-up mean 4.9 years)
239
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
2
none
Stroke (black people) (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious3
none
Stroke (non-black people) (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
very serious
4
none
Combined CVD (black people) (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious3
none
Combined CVD (non-black people) (follow-up mean 4.9 years)
Hypertension (partial update)
Pharmacological interventions
Table 81: Evidence profile comparing ACEi vs CCB in black people and non-black people (data from Leenan et al., 2006)354
NOTE: there was not enough data given in the study to calculate the HRs for these outcomes, so the RRs reported in the paper have been used
in the GRADE profile
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
3
none
data not given in
study
1.03 (0.96,
1.10)
not enough
data given in
study to
calculate
MODERATE
data not given in
study
0.89 (0.75,
1.06)
not enough
data given in
study to
calculate
MODERATE
data not given in
study
0.85 (0.75,
0.97)6
not enough
data given in
study to
calculate
MODERATE
Heart Failure (black people) (follow-up mean 4.9 years)
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious2
none
Heart Failure (non-black people) (follow-up mean 4.9 years)
1
1
randomised
trials
no serious
limitations1
no serious
inconsistency
no serious
indirectness
serious
3
none
Subgroup analysis of RCT: but pre-spcified and the trial deliberately recruited a specific number of black people to be able to do this anlysis
95% confidence interval includes both 1) no effect and 2) appreciable benefit or appreciable harm
3
95% confidence interval excludes no effect, but the CI includes appreciable benefit and non-appreciable benefit or appreciable harm and non-appreciable harm
4
95% confidence interval crosses both 1) no effect and 2) appreciable benefit or harm and non-appreciable benefit or harm
5
SS - favours CCB (p-value not given)
6
SS - favours ACEi (p-value not given)
2
Hypertension (partial update)
Pharmacological interventions
1
randomised
trials
240
Hypertension (partial update)
Pharmacological interventions
10.7.3.2
Economic evidence
No relevant economic studies were identified.
10.7.3.3
Evidence statements
10.7.4
10.7.5
 No relevant cost-effectiveness evidence was identified.
Chronic kidney disease
For guidance pertaining to people with hypertension and chronic kidney disease refer to NICE
Clinical Guideline 73.
Type 1 and Type 2 diabetes
For guidance pertaining to people with hypertension and Type 1 diabetes refer to NICE
Clinical Guideline 15.
For guidance pertaining to people with hypertension and Type 2 diabetes refer to NICE
Clinical Guideline 66.
241
Update 2011
One RCT (subgroup analysis)492 found that:
 Over half (55%) of people who developed angioedema were black people
 The incidence of angioedema (out of all the people who developed angioedema in the trial)
was:
o in black people: higher in the ACEi group versus other a-HT classes (TD, CCB or
alpha) combined (62% vs. 38%)
o in non-black people: lower in the ACEi group versus other a-HT classes (TD, CCB or
alpha) combined (38% vs. 63%)
[moderate quality evidence]
The risk of angioedema in both black people and non-black people was:
 significantly higher in the ACEi group vs. other a-HT classes (TD, CCB or alpha)
combined (as a proportion of the total randomised, see the forest plot in section H.1.4 )
[high and moderate quality evidence]
One RCT (subgroup analysis)354 found that:
 In black people:
 CCB was significantly better than ACEi for risk of:
 Combined CVD
[moderate quality evidence]
 Stroke
[moderate quality evidence]
 There was NS difference between ACEi and CCB for risk of:
 CHD
[high quality evidence]
 HF
[moderate quality evidence]
 In non-black people:
 ACEi was significantly better than CCB for risk of:
 HF
[moderate quality evidence]
There was NS difference between ACEi and CCB for risk of:
 CHD
[moderate quality evidence]
 Combined CVD
[moderate quality evidence]
 Stroke
[low quality evidence]
Hypertension (partial update)
Pharmacological interventions
10.7.6
Women who are pregnant or breast-feeding
For guidance on women who are pregnant or breast-feeding, refer to NICE Clinical Guideline
107 http://guidance.nice.org.uk/CG107.
10.8 Stopping treatment
If a patient's blood pressure has been reduced to normal levels by antihypertensive drugs, both
patient and doctor may want to know if medication can safely be stopped. Unnecessary drug
treatment may put the patient at risk of adverse side effects and is a cost to society. Some
patients may be at risk of serious cardiovascular events if they stop taking antihypertensive
drugs. It would be useful to be able to identify patients who are likely to be able to stop
medication without serious consequences.
In studies which have reported on withdrawal of antihypertensive medication240,349,411,561,631,
421 9,38,201,359,413,433,435,582,597
,
, between 10%433 and 60%349 of patients remained normotensive
for at least a year, although studies reporting better success rates were often of highly selected
patient populations. Further, the definition of normotension varied between studies, from
blood pressure less than 140/85mmHg38 to diastolic blood pressure less than 105mmHg411 and
the characteristics of the patients varied, e.g. mean age ranged from 519,411 to 67 years631,
baseline blood pressure ranged from 126/80 mmHg240,349 to 152/101mmHg359, number of
drugs ranged from one9,201,561,631 to three or more349.
There is consistent evidence, from a systematic review of 5,479 patients who stopped taking
anti-hypertensive medication and who were followed up for at least a year434, and from a
subsequent study of 503 patients who were also followed up for a year435, that patients are
more likely to remain normotensive if they are younger, have lower blood pressure and have
been treated with only one drug. Two studies, of 1,478 patients aged 60–84 years, found that
on-treatment systolic blood pressure was the best measure of blood pressure to use in
predicting success201,435.
We identified three randomised controlled trials of interventions - weight loss and restriction
of salt and alcohol - which might help patients to successfully stop taking anti-hypertensive
medication 349,561,631. The TONE631 and DISH349 studies were similar: they both evaluated the
effects of a weight loss diet and restriction of salt; both randomised obese and non-obese
patients independently; both had weekly group counselling sessions during the initial
intensive phase of the intervention, followed by less frequent group sessions and
individualised counselling during the later maintenance phase; patients in both studies had
good blood pressure control (mean baseline blood pressure 129/72 mmHg in TONE and
127/80 mmHg in DISH). The TONE study enrolled patients who had been taking only one
antihypertensive drug or a combination of a diuretic and a non-diuretic for a mean duration of
11.7 years. The DISH study enrolled patients who had been on treatment for at least 5 years
and included some who were taking three or more antihypertensive drugs. The definitions of
normotension - less than 150/90 mmHg in TONE and diastolic blood pressure less than 95
mmHg in DISH - might now be considered high. Meta-analysis of the results of these trials
showed that obese patients who were put on a diet to lose weight were more likely to be
successful in stopping medication than those who were not (RR = 1.6, 95%CI: 1.4 – 2.0).
Likewise, patients who were encouraged to restrict their salt intake were more likely to
remain normotensive (RR=1.4, 95%CI: 1.2 – 1.7), with little difference between obese and
non-obese patients (see Figure 19). The smaller study by Stamler et al. compared the effects
of a multiple intervention, which encouraged loss of weight and restriction of salt and alcohol,
with no intervention to support drug withdrawal; it defined normotension as diastolic blood
pressure less than 90 mmHg561. This study was combined in a meta-analysis with a similar
comparison of two arms of the TONE study of obese patients: a comparison of the
combination of weight loss and salt restriction with no intervention. Patients who received a
multi-factorial intervention were more likely to successfully stop medication than those who
242
Hypertension (partial update)
Pharmacological interventions
were not (RR = 2.8, 95%CI: 1.9 – 4.0) and these interventions appeared to be more successful
than those which addressed only diet or only salt restriction (see Figure 31). Combining all
groups in these three studies349,561,631, 42% of patients who received interventions remained
normotensive for at least a year, compared to only 25% in the control groups. This is
consistent with the evidence (see Lifestyle interventions) that a healthy diet and reduced salt
intake can lower blood pressure.
Figure 19:
Meta-analysis of RCTs of lifestyle interventions to support withdrawal
of anti-hypertensive drugs
We found little evidence about whether patients became more likely to suffer severe
cardiovascular events if antihypertensive medication was withdrawn. One study monitored
cardiovascular events for 12–32 (average 24) months after withdrawal of medication from
975 patients who had a mean blood pressure of 129/72 mmHg while on one antihypertensive
medication336. It found no difference between the rate of cardiovascular events before and
after withdrawal of medication, though the statistical power to detect a difference was low,
largely because of the short period of monitoring while on medication. The best evidence on
the possible effects of drug withdrawal is the epidemiological evidence from over a million
adults, that any increase in blood pressure is associated with an increased risk of death from
cardiovascular disease361.
If patients become hypertensive after stopping drugs, this is most likely to happen in the first
six months, although it can happen later434. To avoid this, patients should be carefully
followed up and drugs should be withdrawn gradually following manufacturers' guidance.
243
Hypertension (partial update)
Pharmacological interventions
10.9 Link from evidence to recommendations- Pharmacological treatment of
hypertension
244
Update 2011
The pharmacological update of this guideline in 2006 recommended a stepped care approach
to treatment. The recommendation for initial treatment (step 1) was stratified by age and
ethnicity reflecting data from clinical trials showing differential effects of the different classes
of blood pressure lowering drugs on blood pressure lowering and clinical outcomes in
younger (<55years) versus older people and in black people of African and Caribbean
descent. Antihypertensive therapies were designated “A” drugs (ACEi or ARBs), “C” drugs
(calcium channel blockers) and “D” drugs (thiazide-type diuretics). The recommendation for
step 1 treatment for younger people was an “A” drug. At that timethe GDG felt that the
benefit from ACEi and ARBs were closely correlated (although lacked head to head evidence)
and that they should be treated as equal in terms of efficacy; however, due to cost differences,
felt ACE inhibitors should be initiated first and an ARB considered an alternative for when an
ACEi was poorly tolerated, usually due to an ACE-inhibitor-induced cough.
ACE-inhibitors versus ARBs for step 1 treatment:
For this update, the GDG considered evidence from 3 RCTS published since December 2005
comparing ACEi versus ARB for step 1 treatment for adults with primary hypertension. The
first RCT653 (the ONTARGET trial) compared treatment with the ACEi ramipril (10 mg/day)
versus the ARB telmisartan (80 mg/day) and versus a combination of the two (ACEi+ARB) in
25,620 people considered to be at high cardiovascular disease risk. Many (approximately
70%), but not all of these patients had treated hypertension. The study had a median followup time of 56 months. A second RCT587 compared treatment with the ACEi enalapril (20
mg/day) versus the ARB losartan (50 mg/day) in N=560 people with hypertension, for a
follow-up time of 24 months. The third study552 (CORD IB trial) compared treatment with the
ACEi ramipril (5 mg/day) versus the ARB losartan (50 mg/day) in N=3860 people with
hypertension, and had a follow-up time of 12 months. The evidence showed no significant
differences between ACEi and ARBs on major clinical outcomes including death,
cardiovascular events, stroke and diabetes. There was no consistent trend favouring one drug
class over the other. Study drug withdrawal was significantly lower with ARB compared with
ACEi. The GDG considered that this most likely reflected better tolerability of the ARB as
ACEis are known to cause cough in some patients whereas ARBs do not. There was
heterogeneity in the analysis for this latter finding but the lower withdrawal from ARB
therapy was a robust finding in the largest trial (ONTARGET). Moreover, the GDG noted that
there was an eight week run-in to ONTARGET when patients were prescribed the ACEi to
see if they could tolerate the drug, thus, pre-selecting a group with short-term tolerability of
the drugs. The results are therefore likely to underestimate the true withdrawal rate from
ACEi. The GDG noted that side-effects of a drug are an important consideration in making
treatment decisions for the management of a symptomless condition.
The ONTARGET study also compared the combination of ACEi + ARB versus ACEi alone
and found that there was no advantage of the ACEi + ARB combination on clinical outcomes
and a more adverse effects associated with the combination of ACEi + ARB. The GDG
concluded that there was no evidence to support the use of ACEi + ARB for the treatment of
hypertension and that this combination should not be used for the treatment of primary
hypertension.
The largest study in the analysis comparing ACEi versus ARB was ONTARGET and the
GDG discussed the fact that this study was not a trial designed to specifically examine the
treatment of hypertension with initial therapy, but rather looked at the use of an ACEi or ARB
for prevention of cardiovascular events. In this regard, the participants in ONTARGET were
selected to be at high cardiovascular risk, although 70% of patients in ONTARGET had a
history of hypertension and were receiving antihypertensive therapy/s or had discontinued
Hypertension (partial update)
Pharmacological interventions
245
Update 2011
their treatment prior to randomisation to the study drugs. The GDG debated whether ACEi
and ARBs could be considered equivalent, based on data primarily from one large study that
was not specifically a hypertensive population. It was noted that ONTARGET was designed
to test non-inferiority of the ARB versus the most commonly used ACEi (Ramipril) with
regard to clinical outcomes and that further large trials addressing the same question are
unlikely to happen - this may, therefore, be the best evidence ever available for a hypertensive
population. It was reassuring that the other studies in the analysis, albeit much smaller but
studying a more typical hypertensive population, were consistent with the findings of
ONTARGET.
No relevant cost effectiveness analyses comparing ACEi versus ARBs were identified.
However, the difference between the lowest cost ARB and the lowest cost ACEi has reduced
considerably due to the recent availability of generic losartan; generic losartan (100mg) is
now only about £5 more per year than generic ramipril (10mg). Patent expiry is imminent for
many other ARBs too and the GDG considered it likely that the cost of ACEi and ARBs are
likely to become similar over the lifetime of this guideline update.
The ethnicity of participants was not reported for all of the trials but the GDG did not
consider this prevented extrapolation of the findings to a UK population. Finally, the GDG
could not identify any quality of life data comparing ACEi versus ARBs.
The GDG concluded that the drug classes ACE iand ARBs should be considered equivalent
with regard to their effect on clinical outcomes and recommended that people aged <55 years
should be offered step one treatment with an ACEi or a low cost ARB. For patients intolerant
of ACEi, an ARB should be offered. The GDG also recommended that an ACEi and an ARB
should not be combined for the treatment of hypertension. The GDG noted that in women
aged <55years and of child bearing potential, the use of ACEi or ARB has been reported to
increase the risk of foetal malformation if taken during pregnancy. Women taking these
medications should be advised that if they become pregnant, they should discontinue
treatment and inform their doctor. In women planning conception, ACEi and ARBs should be
avoided during this time and alternative treatments considered if required – see clinical
Clinical Guideline 97 on Hypertension in Pregnancy.
Choice of thiazide-type diuretic therapy for hypertension:
The 2006 pharmacological update recommended thiazide-type diuretics as a step 1 treatment
option for people aged ≥55 years or black people of African and Caribbean descent of any age
– the other step 1 option for this group of people being a CCB. There are many different drugs
labelled as thiazide-type diuretics. The predominant thiazide-type diuretic used in the UK for
the treatment of hypertension is low dose (2.5mg o.d.) bendroflumethiazide (BFZ). This is
somewhat unusual because this thiazide-type diuretic is rarely used anywhere else in the
world as the preferred diuretic for the treatment of hypertension. This may be unimportant if
the clinical outcomes data with low dose BFZ is equivalent to that with the other, more
commonly used thiazide-type diuretics elsewhere in the world.
This issue of comparability of different thiazide-type diuretics has been brought into sharper
focus by recognition of the fact that, although often grouped together as thiazide-type
diuretics, from a pharmacological perspective, there are two broad groups; i) classical thiazide
diuretics (e.g. BFZ and hydrochlorthiazide; HCTZ) i.e. the name ends in thiazide, and ii)
thiazide-like diuretics (e.g. chlorthalidone; CTD and indapamide; IND). The thiazide-like
diuretics retain the main action of thiazide diuretics, i.e. inhibition of the sodium chloride cotransporter in the distal nephrons of the kidney. However, the thiazide and thiazide-like drugs
have differential effects on other enzyme effects in the kidney, e.g. carbonic anhydrase
inhibition, which can differ by up to 10,000-fold. Differential effects on platelet aggregation
and regulation of angiogenesis have also been reported. The relevance of these actions beyond
the characteristic thiazide action of inhibition of the sodium chloride cotransporter with regard
to blood pressure control and the prevention of clinical outcomes is unknown. Nevertheless,
Hypertension (partial update)
Pharmacological interventions
246
Update 2011
the GDG considered it important to examine the evidence base supporting the use of classical
thiazides (BFZ or HCTZ) when compared to the thiazide-like diuretics such as CTD and IND.
Another important element of the data review for thiazide-type diuretic therapy was to
examine the doses of diuretics used in the various clinical outcome trials. The trials evaluating
clinical outcomes with thiazide-type diuretics have usually been evaluated by grouping all of
these various drugs used at various doses altogether. The early diuretic trials used much
higher doses than commonly used today. The reduction in dose to what is now known as “low
dose” diuretic therapy resulted from concern about the development of electrolyte
disturbances (usually hypokalaemia) and metabolic disturbances (hyperglycaemia) with
higher dose diuretic therapy. Consequently, the GDG reviewed the important question as to
what is the most clinically and cost effective thiazide-type diuretic for the treatment of adults
with primary hypertension?
The analysis examined data for the four most commonly used thiazide-type diuretics; i)
classical thiazide diuretics (e.g. Bendroflumethiazide (BDZ) and hydrochlorthiazide(HCTZ),
and ii) thiazide-like diuretics (e.g. chlorthalidone (CTD) and indapamide (IND). The analysis
was complex and the GDG noted that there were no direct comparisons between the different
diuretics with regard to clinical outcomes. Where head-to-head comparisons had been
undertaken, they were usually based on blood pressure changes as the main outcome. These
studies were often of short duration and too small to provide robust data. The GDG
considered all of them to be underpowered to detect a significant blood pressure difference
between diuretic treatments. There was also considerable variation in the doses of diuretics
used in the various studies – some early studies using four times the doses used routinely in
today’s clinical practice making it impossible to pool data for analysis. Consequently, the
GDG found it difficult to reach firm conclusions regarding the comparative efficacy of
different thiazide-type diuretics with regard to blood pressure lowering.
The GDG then reviewed the clinical outcome studies with thiazide-type diuretics and found
no direct comparator studies between different diuretics. Furthermore, interpretation of data
from head-to-head trials comparing diuretics with placebo or other antihypertensive drugs
was complicated by the markedly different diuretic doses used across studies. The GDG noted
that the data demonstrating benefits of BFZ on clinical outcomes came from older studies
(MRC) in which the dose of BFZ (10mg o.d.) was four times the usual dose of BFZ i.e. 2.5mg
o.d., used in clinical practice today. The GDG also noted that there was no study evaluating
and confirming the benefit of low dose BFZ on clinical outcomes – the only data coming from
older studies with much higher doses of BFZ, i.e. 10mg od. This concerned the GDG, mindful
of the fact that low dose BFZ (2.5mg o.d.) has been the preferred thiazide-type diuretic for the
treatment of hypertension in the UK. The GDG also noted that there was limited evidence
confirming benefit of initial therapy on clinical outcomes with low doses of
hydrochlorthiazide (12.5-25mg o.d.), the other commonly used thiazide-type diuretic worldwide.
The GDG next discussed the evidence for the thiazide-like diuretics, i.e. IND or CTD and
noted that the there was evidence showing benefits of low dose IND or low dose CTD on a
range of clinical outcomes. The GDG noted that the evidence for IND and CTD was derived
from more contemporary studies that had more consistently used lower doses across studies,
typically; IND 1.5mg SR or 2.5mg o.d., or CTD 12.5mg or 25mg o.d. Some of the IND
studies used an SR formulation, others did not. The GDG concluded that the consistency of
the data suggested that the SR formulation was unlikely to have influenced the clinical
outcomes in studies with IND.
No relevant cost-effectiveness studies were found that compared different types of diuretic.
Current UK drugs costs were considered by GDG and it was noted that the aforementioned
thiazide-type diuretics were all available as generics.
Hypertension (partial update)
Pharmacological interventions
247
Update 2011
Considering all of the data cited above, the GDG were concerned that there was no evidence
confirming a beneficial effect of low dose bendroflumethiazide, i.e. 2.5mg o.d., on clinical
outcomes in people with hypertension. This observation is important because
bendroflumethiazide 2.5mg od. is the most commonly used thiazide-type diuretic for the
treatment of hypertension in the U.K. This does not mean that bendroflumethiazide 2.5mg
o.d. is ineffective but it does make it difficult to assess whether it is as effective at preventing
clinical outcomes as other thiazide-like diuretics, e.g. chlortalidone and indapamide for which
evidence confirming benefits on clinical outcomes does exist. Having undertaken this analysis
it was difficult for the GDG to recommend treatment with low dose thiazide-type diuretics,
e.g. bendroflumethiazide or hydrochlorthoazide for which there was no evidence of a benefit
on clinical outcomes.
Consequently, the GDG recommended that when thiazide-type diuretics are used for the
treatment for primary hypertension, thiazide-like diuretics, e.g. chlortalidone (12.5mg -25mg
od) or indapamide (1.5mg SR or 2.5mg o.d.) should be preferred to conventional thiazide
diuretics, e.g. bendroflumethiazide or hydrochlorthiazide. The GDG did not consider it
necessary to recommend that those people already treated with low dose BFZ and in whom
blood pressure is controlled, should be switched to CTD or IND. However, when new diuretic
therapy was to be initiated, then CTD or IND should be preferred.
The cost-effectiveness of pharmacological treatment of hypertension:
As part of the 2006 pharmacological update of this guideline (CG34), the cost effectiveness of
different classes of antihypertensive medications as initial therapy for hypertension was
evaluated. The analysis assessed the costs and effects of the major antihypertensive drug
classes; (A), i.e. ACE-I / ARB, (B) beta blockers, (C) CCBs and (D) thiazide-type diuretics.
No intervention (NI) was also included as a comparator. Details of this analysis are shown in
appendix x.
Since 2006 the cost of antihypertensive drugs has decreased; in particular the cost of CCBs
and ARBs. The GDG decided that it would be informative to rerun the cost-effectiveness
analysis as part of the 2011 update with updated costs. The base case analysis modelled the
results for 65-year-old men and women with 2% CVD risk, 1% HF risk and 1.1% diabetes
risk. Sensitivity analysis undertaken in 2006 were also rerun to evaluate whether and how the
results varied by age, sex, and by varying the risks of CVD, HF and diabetes. The GDG noted
that the clinical trial evidence on which the model is based included relatively few younger
(under 55) people, so speculative sensitivity analyses were conducted to explore how different
assumptions about treatment effects might impact on the cost-effectiveness results for
younger (under 45) people.
The top line conclusion from this analysis is that treating hypertension is highly costeffective. Treatment resulted in improved health outcomes (higher QALYs) and remarkably,
with most of the drug classes in the model, actually resulted in overall cost savings when
compared to no treatment. This cost saving is due to the fact that the reduction in
cardiovascular events led to savings that offset the relatively low cost of antihypertensive
medication. The GDG noted that this conclusion is based on the use of low cost generic drugs.
Another important conclusion is that for most people, CCBs were found to be the most costeffective treatment option for initial treatment of primary hypertension. Indeed, unlike the
analysis in 2006, CCBs are now cost saving when compared to no intervention.
The GDG noted another key difference from the 2006 analysis is that the absolute difference
in costs between ACE/ARB, CCBs and thiazide-type diuretics is now much smaller than it
was in 2006. The difference is QALYs between these drugs is also fairly small. Just as in
2006, beta-blockers are ruled out by simple dominance, however now all other treatments are
estimated to be both cheaper and more effective – further justifying the decision not to
recommend beta-blockers as a preferred initial therapy for primary hypertension.
The GDG then reviewed the cost-effectiveness analysis in various sub-groups and noted that
when compared to the 2006 analysis, CCBs are most cost effective in a greater number of
scenarios. The GDG noted that the sub-group analysis of cost-effectiveness was particularly
sensitive to the relative effects of drug therapy on the prevention of diabetes and heart failure.
The model predicts that for people at low to intermediate risk of heart failure, CCBs are the
most cost-effective option because they are associated with a low risk of developing diabetes,
especially when compared to thiazide type diuretics, and they also have a good effectiveness
profile across the range of other CVD risks.
Conversely, when people are judged to be at a high risk of developing heart failure, thiazidetype diuretics were estimated to be the most cost-effective option, provided that they do not
also have a high risk of diabetes. For people with a high risk of both heart failure and
diabetes, ACE inhibitors or ARBs may be the most cost-effective option. The GDG noted that
the applicability of this data to people under the age of 55 is uncertain, since it is based on
trial data from mostly older people. Furthermore, although the model was robust to a variety
of sensitivity analyses, there remains uncertainty about the size of some treatment effects,
which translates into uncertainty about the relative cost-effectiveness of the drugs.
The GDG considered the implications of the cost-effectiveness analysis with regard to the
preferred treatment strategy for hypertension. Most people with primary hypertension are a
low-to intermediate risk of heart failure and have an increased risk of developing diabetes,
this suggests that CCBs would be the most cost-effective step 1 therapy for most people aged
over 55 years. The caveat to this conclusion is that the risk of heart failure increases with
increasing age, especially in the elderly (i.e. ≥80 years) in whom a thiazide-like diuretic
would be a more cost effective treatment. Moreover, some people might not tolerate a CCB or
may have evidence of oedema that might benefit from the preferred used of a thiazide-type
diuretic.
The GDG concluded that the cost-effectiveness analysis demonstrated that CCBs are the most
cost-effective initial therapy for most people aged >55 years with primary hypertension, and
indeed, cost saving when compared to no intervention. It was considered that the evidence
supporting this conclusion was stronger than in 2006. In addition the GDG discussions around
this recommendation highlighted new data demonstrating; i) that CCBs appear to be the most
effective treatment option to suppress blood pressure variability, which in turn appears to be
an independent predictor of cardiovascular disease risk in people with treated hypertension
(see below); and ii) that new evidence suggests that for treatment at step 2, the combination
of A + C will usually be preferred to A + D, thereby impacting on the preferred choice of
therapy for step 1 treatment (see section below – step 2 treatment). Consequently, the GDG
recommended that a CCB should be the preferred initial therapy for people with primary
hypertension and aged >55 years. A thiazide-like diuretic (i.e. chlortalidone or inadapamide)
are considered a suitable alternative for those who cannot tolerate a CCB or who have
developed, or are at high risk of developing heart failure.
Blood Pressure Variability and the impact of Antihypertensive therapy:
Just after the scope for this guideline update had been finalised, a series of analyses were
published showing that excessive variability in blood pressure is an independent risk factor
for cardiovascular events, over and above the effect of the level of blood pressure itself.
Furthermore, a systematic review of previous trials suggested that different classes of
antihypertensive medications varied in their capacity to influence blood pressure variability.
The GDG decided to review this data as part of this update (see Appendix F.1). The GDG
noted that blood pressure variability can be measured in a number of ways but is perhaps most
easily understood when expressed at the standard deviation (SD) around the mean of a
number of blood pressure readings. The series of blood pressure readings may have been
taken repeatedly at a single clinic visit, or an analysis of the variation between clinic visits, or
across a series of measurements recorded by ABPM. Put simply, two people could have the
248
Update 2011
Hypertension (partial update)
Pharmacological interventions
011
Hypertension (partial update)
Pharmacological interventions
same mean blood pressure but a different SD value for multiple readings, reflecting
differences in blood pressure variability. This can be expressed as systolic or diastolic
pressure variability. The studies reviewed by the GDG involved a series of retrospective
analyses of clinical trial data (see appendix x). Review or these studies showed that variability
in systolic blood pressure when measured visit-to-visit was a strong predictor of stroke,
independent of mean systolic blood pressure. Moreover, in people with treated hypertension,
a higher residual blood pressure variability is associated with a higher risk of vascular events.
The GDG noted that it was unclear if blood pressure variability was causally related to
clinical outcomes, or a marker of more severe underlying vascular disease. Furthermore,
blood pressure is highly variable and although less so when measured under standardised
conditions, it is unclear what the boundaries of normal versus abnormal variability would be
in usual clinical practice. The GDG agreed that whatever the underlying mechanisms, systolic
blood pressure variability appears to be an important independent predictor of clinical
outcomes.
The GDG also reviewed data from a systematic review and meta-analysis which examined the
effect of different classes of blood pressure treatment on blood pressure variability in trials.
This analysis revealed that blood pressure variability was most effectively reduced by CCBs,
closely followed by thiazide-type diuretics. The analysis also showed that beta-blockers were
the least effective and may actually increase blood pressure variability.
Having considered these findings on blood pressure variability the GDG concluded that those
most at risk of having increased systolic blood pressure variability, i.e. older hypertensive
people, will already be treated with the most effective drug classes to suppress systolic blood
pressure variability, i.e. a CCB (or a thiazide-like diuretic if a CCB is not indicated or
tolerated) as step 1 therapy, according to the recommendations in this guideline update. The
GDG concluded that the updated guidance recommends the best available evidence-based
treatment options to suppress blood pressure variability in people with hypertension.
Step two therapy:
Many people with treated hypertension will require more than one drug to control their blood
pressure. For people whose blood pressure is not controlled by step 1 treatment, i.e. A in
younger adults (≤55years) or C or D in people aged >55yrs, the 2006 pharmacological
update of this guideline recommended that step 2 therapy should be a combination of A + C
or A + D. the choice of which combination was solely dictated by whether the patient was
commenced on treatment with C or D at step 1. This reflected the fact that at the time of the
2006 update, there was no published data to better inform the discussion about whether there
was a preferred combination for most people at step 2.
For this 2011 update of the guideline, one RCT 296 was found which prospectively examined
the effect of A + C versus A + D on clinical outcomes in the ACCOMPLISH trial. This study
compared treatment with the ACE-i benazepril + the CCB amlodipine vs. the ACE-i
benazepril + the thiazide diuretic hydrochlorothiazide in 11,506 people with hypertension, for
a follow-up of 24 months.
The GDG discussed the evidence which showed that ACE+CCB was significantly more
effective at preventing MI when compared to ACEi + diuretic. Study withdrawal was also
significantly lower in patients randomised to treatment with the combination of ACEi+CCB.
The other clinical outcomes were not significantly different between groups but all
numerically favoured the ACEi + CCB combination. The GDG noted that the ACCOMPLISH
trial was stopped earlier than planned because the primary composite outcome was
significantly in favour of the ACEi + CCB. Thus, the study had inadequate power to address
individual cardiovascular outcomes. There was no quality of life data identified.
The GDG concluded that the combination of ACEi+CCB had a treatment advantage over
ACEi+diuretic. However, the GDG noted that this conclusion is based on a single large study.
The GDG also noted that the ACEi used in this study, i.e.benazepril, is not used in the UK but
249
Hypertension (partial update)
Pharmacological interventions
250
Update 2011
concluded that there was unlikely to be an important difference between benazepril and other
ACEi. Likewise, the GDG considered it likely that the results with the ACEi + CCB would be
replicated with an ARB + CCB. The GDG also considered the black people of African or
Caribean origin, ACEi are associated with an increased risk of developing angioedema which
can be life threatening. Although the incidence of this adverse of ACEi in back people of
African or Caribean origin is low, the GDG suggested that an ARB in preference to an ACEi
should be considered for such patients when step 2 treatment in required. The GDG
concluded that this data from the ACCOMPLISH trial, taken together with the updated costeffectiveness analysis and the data on blood pressure variability, all favour the combination of
A + C versus A +D – with the caveat that the differences between C and D in each of these
areas of analysis, whilst usually favouring C, was not large. The GDG emphasised that whilst
a CCB should usually be preferred versus thiazide-like diuretic as step 1 and step 2 therapy
for most people, a thiazide-like diuretic is a highly effective alternative and is preferred in
people with evidence or, or at high risk of developing heart failure.
The GDG recommended that A + C should be the preferred step 2 therapy for most patients.
A+D is an alternative step 2 treatment in those intolerant of a CCB or in those with a high risk
of heart failure.
Step 3 Treatment for Hypertension:
The GDG did not formally review new evidence for step 3 treatment for the 2011 update.
However, the GDG discussed the implications of the recommendations for step 1 and 2
treatments with regard to step 3 treatment. The GDG concluded that it follows from the
evidence reviews cited above that the recommended step 3 treatment should be; A (ACEi or
ARB) + CCB + D (thiazide-like diuretic, i.e. chlothalidone or indapamide).
Resistant hypertension: (step 4 treatment)
The GDG decided that the term ‘resistant hypertension’ should be applied to people requiring
step 4 treatment and defined resistant hypertension as follows;
Definition of Resistant Hypertension: A person with resistant hypertension is someone who
has confirmed hypertension and in whom clinic blood pressure is not controlled
(<140/90mmHg) despite treatment with a rational combination of optimum or best tolerated
doses of three antihypertensive drugs (usually A+C+D).
The GDG noted that poor compliance with therapy and white coat hypertension could each
manifest as apparent resistance to drug treatment and should be considered. Secondary causes
for hypertension should also be reconsidered in people with resistant hypertension and
discussion with a specialist may be required to address some of these issues.
Based on health survey for England data, the GDG estimated that resistant hypertension is
likely to affect approximately 500,000 people with treated hypertension in the U.K. and thus
represents an important clinical problem. These people will be older and often have
established cardiovascular disease, diabetes or CKD and thus, be at high cardiovascular risk.
From a cardiovascular risk perspective, such people potentially have much to gain in terms of
absolute benefit from further blood pressure lowering.
The GDG noted that the treatment of resistant hypertension has not been studied in detail, in
part because few drugs are developed that are specifically targeted at resistant hypertension.
There is as a consequence, a paucity of data upon which to base guidance for the treatment of
resistant hypertension. For the 2006 pharmacological update of this guideline, there was no
formal evidence review for step 4 treatment and the GDG cautiously recommended a range of
options that included; “further diuretic therapy”, alpha blockers or beta blockers. For this
2011 update the literature was searched for all years and all study types were included.
Populations which were exclusively diabetic or had chronic kidney disease were excluded.
The data search failed to indentify a single head-to-head RCT that met our search criteria. Six
studies did meet the search criteria, however, these were all retrospective cohort studies – i.e.
post-hoc analyses of studies in which patients had been treated with four or more
Hypertension (partial update)
Pharmacological interventions
251
Update 2011
antihypertensive therapies. The GDG noted that all of these studies evaluated the use of low
doses of spironolactone (an aldosterone antagonist), usually 25mg o.d. Together, the review
of this data suggested that low dose spironolactone was effective in resistant hypertension
based on the surrogate outcome of blood pressure lowering. There was no data on other
clinical outcomes. It is unclear from this very limited data whether spironolactone is always
the most effective treatment option for every patient with resistant hypertension. Furthermore,
the GDG noted that spironolactone is not licensed for the treatment of hypertension in the
U.K. but this does not preclude its use. Not all people are able to tolerate spironolactone, the
main adverse effect being the development of nipple tenderness and/or gynaecomastia in
males. Another important consideration is that spironolactone is a potassium sparing diuretic
and may cause hyperkalaemia, especially when combined with an ACE-inhibitor or ARB, as
will be the case for most people with resistant hypertension treated according to the algorithm
recommended by this guideline. The GDG considered this to be a very important safety
issue. Where reported, the studies that have used spironolactone for the treatment of resistant
hypertension have not used it when the baseline potassium level exceeded 5.00mol/L, and
spironolcatone was used with caution in people with a reduced eGFR. The GDG discussed
these safety aspects and recommended that in primary care, low dose spironolactone should
only be considered for the treatment of resistant hypertension when the blood potassium level
is <4.5mmol/L. Particular caution is advised in people with a reduced GFR as they are at
increased risk of hyperkaelemia and renal function should be monitored closely in all patients
receiving sprinolactone. Blood potassium, sodium and creatinine values should be checked
approximately 2 weeks after treatment initiation and perdiodically thereafter.
The GDG also highlighted that patients should be advised to discontinue spironolactone
treatment if they become significantly dehydrated due to illness such as vomiting and/or
diaorrhea. The GDG recognised that the emphasis of too many caveats and concerns might
limit the use of what can be a very effective drug in the setting of resistant hypertension.
Nevertheless, care is needed to monitor patients when treatment regimens become
increasingly complex.
The GDG discussed the potential use of other drug classes for resistant hypertension and
noted that treatments such as higher doses of thiazide type diuretics, alpha blockers and beta
blockers have been used as add-on therapy in clinical trials at step 2 and 3 but not necessarily
at step 4. The GDG concluded that this provides some evidence for the potential effectiveness
of these other treatment options as “add-on” therapy. The GDG also considered alternative
“further diuretic therapy” to spironolactone if this was deemed inappropriate treatment
because of an elevated baseline potassium level or concerns about renal function. The GDG
concluded that If blood potassium levels are higher than 4.5 mmol/l, then higher-dose
thiazide-like diuretic treatment may be considered as an alternative. The GDG also discussed
newer therapies such as the direct renin inhibitor aliskiren but concluded that there was
insufficient evidence of its effectiveness to determine its suitability for use in resistant
hypertension.
In summary, the GDG concluded that resistant hypertension is an important clinical problem
that has been poorly studied with regard to the underlying causes and the most effective
treatment options. Clinicians should consider referral of people with resistant hypertension for
specialist advice/evaluation – especially those who are younger and those with complex
comorbidities. The best evidence, albeit weak evidence, suggests that low dose spironolactone
(e.g. 25mg o.d.), when safe to use and when tolerated, can be an effective means of further
lowering blood pressure. It is unclear if this is the optimal treatment for most people with
resistant hypertension or whether other treatment options would be more effective in most or
some cases. When use of spironolactone is not possible or not tolerated, then higher dose
thiazide-like diuretic, alpha blockers or beta blockers are suitable alternatives for step 4
treatment, with the caveat that the evidence base is very limited and careful monitoring of
Hypertension (partial update)
Pharmacological interventions
electrolytes and renal function is essential. The GDG recognised the need of more research in
this area.
Recommendations
41. Where possible, recommend treatment with drugs taken only once a day. [2004]
42. Prescribe non-proprietary drugs where these are appropriate and minimise cost. [2004]
43. Offer people with isolated systolic hypertension (systolic BP 160 mmHg or more) the
same treatment as people with both raised systolic and diastolic blood pressure. [2004]
45. Offer antihypertensive drug treatment to women of child-bearing potential in line with
the recommendations on Management of pregnancy with chronic hypertension and
Breastfeeding in ‘Hypertension in pregnancy’ (NICE clinical guideline 107). [2010]
Step 1 treatment
46. Offer people aged under 55 years step 1 antihypertensive treatment with an angiotensinconverting enzyme (ACE) inhibitor or a low-cost angiotensin-II receptor blocker (ARB). If
an ACE inhibitor is prescribed and is not tolerated (for example, because of cough), offer a
low-cost ARB.e [new 2011]
47. Do not combine an ACE inhibitor with an ARB to treat hypertension.e [new 2011]
48. Offer step 1 antihypertensive treatment with a calcium-channel blocker (CCB) to people
aged over 55 years and to black people of African or Caribbean family origin of any age. If
a CCB is not suitable, for example because of oedema or intolerance, or if there is evidence
of heart failure or a high risk of heart failure, offer a thiazide-like diuretic. [new 2011]
49. If a diuretic is to be initiated or changed, offer a thiazide-like diuretic, such as
chlortalidone (12.5 mg–25.0 mg once daily) or indapamide (1.5mg slow release or 2.5 mg
once daily) in preference to a conventional thiazide diuretic such as bendroflumethiazide or
hydrochlorothiazide. [new 2011]
50. For people who are already having treatment with bendroflumethiazide or
hydrochlorothiazide and whose blood pressure is stable and well controlled, continue
treatment with the bendroflumethiazide or hydrochlorothiazide. [new 2011]
51. Beta-blockers are not a preferred initial therapy for hypertension. However, beta-blockers
may be considered in younger people, particularly:
 those with an intolerance or contraindication to ACE inhibitors and angiotensin-II
receptor antagonists or
e
In 2007 the MHRA issued a drug safety update on ACE inhibitors and angiotensin II receptor antagonists: not for use in
pregnancy that states ‘Use in women who are planning pregnancy should be avoided unless absolutely necessary, in
which case the potential risks and benefits should be discussed’. There is also a 2009 MHRA safety update for ACE
inhibitors and angiotensin II receptor antagonists: use during breast feeding and related clarification: ACE inhibitors and
angiotensin II receptor antagonists.
252
Update 2011CCC
44. Offer people aged 80 years and over the same antihypertensive drug treatment as people
aged 55–80 years, taking into account any comorbidities. [new 2011]
Hypertension (partial update)
Pharmacological interventions
 women of child-bearing potential or
 people with evidence of increased sympathetic drive. [2006]
52. If therapy is initiated with a beta-blocker and a second drug is required, add a calciumchannel blocker rather than a thiazide-like diuretic to reduce the person’s risk of
developing diabetes. [2006]
Step 2 treatment
53. If blood pressure is not controlled by Step 1 treatment, offer step 2 treatment with a CCB
in combination with either an ACE inhibitor or an ARBef. [new 2011]
55. For black people of African or Caribbean family origin, consider an ARBg in preference
to an ACE inhibitor, in combination with a CCB.e [new 2011]
Step 3 treatment
56. Before considering step 3 treatment, review medication to ensure step 2 treatment is at
optimal or best tolerated doses. [new 2011]
57. If treatment with three drugs is required, the combination of ACE inhibitor or
angiotensin-II receptor blocker, calcium-channel blocker and thiazide-like diuretic should
be used.e [2006]
Step 4 treatment
58. Regard clinic blood pressure that remains higher than 140/90 mmHg after treatment with
the optimal or best tolerated doses of an ACE inhibitor or an ARB plus a CCB plus a
diuretic as resistant hypertension, and consider adding a fourth antihypertensive drug
and/or seeking expert advice.e [new 2011]
59. For treatment of resistant hypertension at step 4:
 Consider further diuretic therapy with low-dose spironolactone (25 mg once daily)h. If
the blood potassium level is 4.5 mmol/l or lower. Use particular caution in people with a
reduced estimated glomerular filtration rate because they have an increased risk of
hyperkaelemia.
 Consider higher-dose thiazide-like diuretic treatment if the blood potassium level is
higher than 4.5 mmol/l. [new 2011]
60. When using further diuretic therapy for resistant hypertension at step 4, monitor blood
sodium and potassium and renal function within 1 month and repeat as required thereafter.
[new 2011]
f
Choose a low-cost ARB.
Choose a low-cost ARB.
h
At the time of publication (August 2011), spironolactone did not have UK marketing authorisation for this indication.
Informed consent should be obtained and documented.
g
253
Update 2011
54. If a CCB is not suitable for step 2 treatment, for example because of oedema or
intolerance, or if there is evidence of heart failure or a high risk of heart failure, offer a
thiazide-like diuretic. [new 2011]
Hypertension (partial update)
Patients’ perspectives
61. If further diuretic therapy for resistant hypertension at step 4 is not tolerated, or is
contraindicated or ineffective, consider an alpha- or beta-blocker. [new 2011]
62. If blood pressure remains uncontrolled with the optimal or maximum tolerated doses of
four drugs, seek expert advice if it has not yet been obtained. [new 2011]
Research recommendations
6. In adults with hypertension, which drug treatment (diuretic therapy versus other step 4
treatments) is the most clinically and cost effective for step 4 treatment?
Although this guideline provides recommendations on the use of further diuretic therapy for
treatment at step 4 (resistant hypertension), they are largely based on post-hoc observational
data from clinical trials. More data are needed to compare further diuretic therapies, for
example a potassium-sparing diuretic with a higher-dose thiazide-like diuretic, and to
compare diuretic therapy with alternative treatment options at step 4 to define whether further
diuretic therapy is the best option.
11
Patients’ perspectives
11.1 Introduction
A published survey that examined the views of 452 hypertensive patients in one urban GP
practice illustrated the range of feelings surrounding the taking of antihypertensive
medications. There was a 77% response rate among patients invited to participate71. Four in
every five people taking part in the study said they had reservations about taking
antihypertensives. Over a third of patients reported experiencing current or previous side
effects from blood pressure lowering medication and nearly 40% were concerned by the
potential harm caused by the long term use of such drugs. Thirty-six percent of responders
wondered if they still needed blood pressure lowering medication and two-thirds would prefer
non-drug therapy. The most commonly cited reasons for taking antihypertensive medications
were 'to achieve some good results' (92%), 'because of what happens at the doctors' (87%) and
'because it feels reassuring' (68%). Before starting on tablets to treat high blood pressure,
patients often weighed the potential benefits against reservations in the context of a personal
framework.
Information available on the DIPEx website (www.dipex.org) was summarised and discussed
by the guideline development group. The DIPEx web site reflects patients' experiences of
serious illness, aiming to share experiences, provide patient friendly information, answer
common questions and provide information on relevant organisations and support groups to
patients, family and friends, carers and health professionals.
The hypertension module contains transcribed interviews from 40–50 people who have
experienced hypertension and can be viewed as transcripts, video or audio clips of
individuals, or collated information on specific topics. The modules are produced by an
advisory panel of patients, health professionals and social scientists with relevant expertise.
Below is a summary of patients' accounts of discovery, treatment and living with
hypertension.
11.2 Discovering hypertension
The route to diagnosis of hypertension was varied, with some patients detected during routine
screening whilst others were identified after a specific event, for example a transient
ischaemic attack (TIA), or following a consultation for a specific problem, for example
dizziness or chest pain. Many patients perceived stress as a major causative factor, even to the
extent that they would blame stresses in their lives of which they had previously been
254
Hypertension (partial update)
Patients’ perspectives
unaware. Other factors which they linked to hypertension were family history, genetic makeup, race, personality traits and specific habits such as alcohol consumption, smoking and salt
intake. Patients reported a degree of frustration when they had eliminated factors they
believed to contribute to their hypertension only to find that their blood pressure remained
unchanged.
Many of those interviewed felt that they had not been given sufficient information regarding
the cause of their hypertension. Attitudes were influenced by patients' background knowledge
about hypertension and whether they were asymptomatic at diagnosis. Some patients
exhibited a positive attitude, feeling that detection gave them the opportunity to modify their
lifestyle and for their hypertension to be monitored and treated to prevent long term disease.
Others felt that their hypertension might have been detected earlier if doctors had been more
vigilant.
11.3 Treatment
Patients voiced a great deal of concern over the issue of long term medication, highlighting
potential side effects and the cost and need for regular prescriptions as major worries. Many
patients reported no problems with antihypertensive drugs, but others had experienced a
variety of side effects. Patients were most concerned about taking beta-blockers and these
were perceived as having a higher side effect profile. ACEi and calcium-channel blockers
were more favoured. Some patients found it difficult to accept side effects of blood pressure
lowering medication when they were asymptomatic. In particular, drugs which led to
impotence were considered unacceptable. Compliance to medication was also an issue, and
many reported that they found it difficult to remember to take tablets. Some patients accepted
that taking tablets was just part of everyday life, whilst others felt it to be a constant reminder
of living with disease. Patients often felt under pressure from family members or health care
professionals to be compliant and selecting the right combination of tablets often led to
anxiety as patients were changed from one medication to another. In attempts to avoid or
delay drug therapy, a proportion of patients wanted to try lifestyle measures or
complementary therapies as an initial alternative to blood pressure lowering drugs.
11.4 Living with hypertension
Many patients were unsure of what it meant to have a diagnosis of hypertension - how serious
was it? The increased risk of stroke and heart disease led some to focus on personal mortality,
and to worry about dependants or financial issues if such events were to occur. Some patients
reported that nothing really changed whilst others now viewed themselves as unhealthy or
even experienced denial.
Patients were anxious as they found it difficult to regulate their behaviour, particularly as they
did not have changing symptoms, so as not to further increase their risks of cardiovascular
disease. Others reported symptoms that they thought were related to hypertension such as
headache, dizziness and visual problems. Often side effects of tablets were attributed to
disease.
Most patients made some attempt to incorporate lifestyle changes, such as restricting salt
intake, increasing exercise and reducing stress. Patients often felt they wanted advice from
health care professionals to avoid 'self-harm' and reported feelings of guilt and frustration if
targets were not achieved. In general, patients welcomed information provided by general
practitioners; some felt doctors did not provide enough information and looked for other
sources such as the web, media or medical magazines. Others felt doctors pitched information
- both the amount and content - at just the right level. A minority of patients felt that the
greater their understanding about high blood pressure, the more that they had to worry about.
Other patients found that people's accounts of living with hypertension were a valuable source
of reassurance; however, they acknowledged that speaking openly about this was often
difficult. Some expressed the view that having hypertension was a very private issue, rarely
255
Hypertension (partial update)
Patients’ perspectives
discussed, but felt that talking did provide much needed support and welcomed sites such as
DIPEx as a forum in which to share their experiences.
11.5 Education and adherence
11.5.1
Compliance with Prescribed Antihypertensive Medication
It is estimated that between 50–80% of patients with hypertension do not take all of their
prescribed medication377,518. This has implications for the successful management of
hypertension with poor adherence to medications linked to inadequately controlled blood
pressure273. Understanding patient's reasons for not taking medications and implementing
effective strategies to overcome barriers to taking prescribed medication is therefore a crucial
aspect in the management of hypertension.
Compliance is used variably as a term within the literature, referring sometimes to the
constant neglect of treatment346, 344 and sometimes to a range of behaviours including delay in
dosing, skipping a dose, longer lapses in dosing and over compliance when extra doses are
taken620. It has been argued that recognizing these differences in compliance patterns is
valuable in working with patients on improving their adherence to prescribed drug
regimens620. Compliance has also been challenged as a concept because of its implied
paternalism and failure to see patients as active, intentional and responsible participants in
their health care management346, 344. Increasingly the term concordance is used within the
literature, implying a more interactive and participatory approach to drug prescribing518.
Not only is it important that drug regimens are adhered to in order to control blood pressure
but it has also been suggested that partial compliance and erratic patterns of dosing may do
more transient harm than any overall beneficial effect of treatment143. For example abrupt
discontinuation of medications may lead to rebound hypertension with elevated blood
pressure. Variability in blood pressure caused by abrupt changes in drug taking patterns has
been linked to certain kinds of target organ damage such as pulmonary congestion and a
consequent deterioration of congestive heart failure143. Therefore strategies to improve
adherence also need to address the need to maintain regular and consistent patterns of drug
usage.
There are many factors that influence patients' decisions not to take their drugs as
prescribed70,267. Factors most pertinent for patients suffering from hypertension include the
asymptomatic nature of the disease. A condition without symptoms combined with the
possibly unpleasant side effects of treatment may contribute to a patient's decision to stop or
reduce their medication83. The long term nature of the treatment is also a factor that can lead
to poorer compliance. Drug complexity, poor instructions, poor provider-patient relationships
and patient's disagreement about their need for treatment may also serve as a reason for nonadherence to drug regimens267.
A wide range of interventions have been developed to try and help patients follow their
prescribed drug regimens. These have included simplified dosing, educational interventions,
telephone and computer assisted monitoring, family interventions, increased convenience of
care with provision of care at the work site, and a team approach with increased involvement
of a community nurse and/or a community pharmacist267,518.
Two systematic reviews have sought to assess the effectiveness of these interventions267,292.
One looked specifically at the relationship between daily dose frequency and adherence to
antihypertensive medication292. In a meta-analysis of data from eight studies it was found that
the average adherence rate was significantly higher for patients with once daily dosing
compared taking those taking multiple daily doses (91% vs. 83%). Adherence rates were also
significantly higher for patients taking once daily doses compared with twice daily doses
(93% vs. 87%). The difference in adherence rates between twice daily and multiple daily
dosing was not significant. Simplifying dosing regimens to once daily use appears to promote
compliance. However it is insufficient on its own to result in adequate compliance and the
256
Hypertension (partial update)
Patients’ perspectives
11.5.2
medical consequences may be graver for patients failing to adhere to once daily regimens,
since missing one dose will result in missing the total daily dose.
A narrative review of a wide range of interventions designed to increase compliance with
prescribed drug regimens across a range of chronic disease entities found that half were
associated with a statistically significant increase in medication adherence but that many were
too small to show an effect. However they concluded that even the most effective
interventions did not lead to large improvements in adherence and treatment outcomes267.
Whilst they may not result in large improvements in adherence to prescribed drug treatments
it would appear that improving patient education, providing counselling, involving families
and other members of the health care team can all have a positive impact. Qualitative research
methods have also contributed to an understanding of how patients weigh up their
reservations about treatment against different reasons for taking treatment: this involves
positive experiences with doctors, perceived benefits of medication and pragmatic
considerations70. Patients will balance reservations and reasons differently. Greater adherence
to drug treatment might be achieved if health care professionals asked patients how they
perceived the advantages and disadvantages of taking medication and listened to their
reservations, their reasons for taking medication and the balance between the two.
Implementing lifestyle measures
Lifestyle interventions such as weight reducing diets, lowering salt intake, exercise, alcohol
reduction and relaxation therapy can reduce blood pressure and it is recommended that
patients are given advice to promote such lifestyle changes. However, it is recognised that
lifestyle changes are difficult to adopt and their effectiveness is often limited. The concept of
compliance has now evolved to encompass 'an active, intentional and responsible process
whereby patients work to maintain their health in collaboration with health care personnel'
rather than simply patients' adherence to instructions344. Many factors are thought to influence
adherence including age, sex, education, understanding and disease perspectives, the mode of
delivering advice and the type of health system647. Adherence may be improved by good
communication between patients and health professionals addressing knowledge about
disease, active involvement of patients in decisions, setting achievable goals and good family
and community support344,358,647.
Adherence with lifestyle modifications, especially dietary changes, is lower than with
antihypertensive drug therapy by between 13% and 76%109. Few studies specifically address
this issue and most research on adherence to lifestyle advice examines strategies to reduce
cardiovascular risk. Important issues to consider are the characteristics of the 'information
provider', the 'information receiver', the 'information itself' and the dissemination strategy.
Who should give it?
In many instances, lifestyle advice is given by nurses who manage clinics for the secondary
prevention of coronary heart disease. These nurse-led initiatives have been shown to be
effective at modifying lifestyle behaviours, reducing blood pressure, monitoring medication
and ultimately in reducing mortality112,417. The regular follow-up provided by these clinics
may help compliance358. The Department of Health has provided guidance for general
practitioners and practice nurses who wish to refer patients to facilities such as leisure centres
or gyms for supervised exercise programmes173.
How should it be given?
Advice alone is less effective than specifically adapted programmes supported by written and
audiovisual material109,605. Material tailored to meet the educational and cultural needs of the
population it is targeting has also been shown to be effective342.
257
Hypertension (partial update)
Reference list
Who should receive it?
Targeting of advice to higher risk populations is thought to be more clinically and cost
effective. A systematic review of 18 trials examining the effects of multiple risk factor
interventions (stopping smoking, exercise, dietary control, weight control, antihypertensive
drugs and cholesterol lowering drugs) in the primary prevention of coronary heart disease in
middle aged adults showed little overall effect on mortality. However, it was noted that
hypertensive 'high risk' patients were more likely to benefit from counselling, education and
effective drugs and thus targeting health education to this group might be of some value186.
11.5.3
What are the most successful strategies for information delivery?
A review of 46 studies on compliance with drug therapy and lifestyle modifications in
cardiovascular risk reduction identified the following effective strategies; behavioural skill
training, self monitoring, telephone/mail contact, self-efficacy enhancement and external
cognitive aids358. A review of compliance with low salt diets suggested that successful
interventions require specific goals, delegation of responsibilities, in-depth patient
assessment, behavioural motivation, implementation plans, repetitive education and extensive
monitoring376. Delivering programmes through specific channels, for example community
based projects may increase effectiveness358.
Recommendations
63. Provide appropriate guidance and materials about the benefits of drugs and the unwanted
side effects sometimes experienced in order to help people make informed choices. [2004]
64. People vary in their attitudes to their hypertension and their experience of treatment. It
may be helpful to provide details of patient organisations that provide useful forums to
share views and information. [2004]
65. Provide an annual review of care to monitor blood pressure, provide people with support
and discuss their lifestyle, symptoms and medication. [2004]
12
Reference list
1
Effects of treatment on morbidity in hypertension. II. Results in patients with diastolic
blood pressure averaging 90 through 114 mm Hg. JAMA. 1970; 213(7):1143-1152.
2
Effect of antihypertensive treatment on stroke recurrence. Hypertension-Stroke
Cooperative Study Group. JAMA. 1974; 229(4):409-418.
3
Evaluation of drug treatment in mild hypertension: VA-NHLBI feasibility trial. Plan
and preliminary results of a two-year feasibility trial for a multicenter intervention study to
evaluate the benefits versus the disadvantages of treating mild hypertension. Prepared for the
Veterans Administration-National Heart, Lung, and Blood Institute Study Group for
258
Update 2011
66. Because evidence supporting interventions to increase adherence is inconclusive, only
use interventions to overcome practical problems associated with non-adherence if a
specific need is identified. Target the intervention to the need. Interventions might include:
 suggesting that patients record their medicine-taking
 encouraging patients to monitor their condition
 simplifying the dosing regimen
 using alternative packaging for the medicine using a multi-compartment medicines
system. (This recommendation is taken from ‘Medicines adherence’, NICE 408 clinical
guideline 76). [new 2011]
Hypertension (partial update)
Reference list
Evaluating Treatment in Mild Hypertension. Annals of the New York Academy of Sciences.
1978; 304:267-292.
4
The Australian therapeutic trial in mild hypertension. Report by the Management
Committee. Lancet. 1980; 1(8181):1261-1267.
5
Hydrochlorothiazide and bendroflumethiazide in low doses--a comparative trial. Acta
Pharmacologica Et Toxicologica. 1984; 54 Suppl 1:47-51.
6
An international trial of antihypertensive therapy in elderly patients. Objectives,
protocol and organization. European Working Party on High Blood Pressure in the Elderly
(EWPHE). Archives Internationales De Pharmacodynamie Et De Therapie. 1985;
275(2):300-334.
7
Cardiovascular risk and risk factors in a randomized trial of treatment based on the
beta-blocker oxprenolol: the International Prospective Primary Prevention Study in
Hypertension (IPPPSH). The IPPPSH Collaborative Group. Journal of Hypertension. 1985;
3(4):379-392.
8
MRC trial of treatment of mild hypertension: principal results. Medical Research
Council Working Party. BMJ. 1985; 291(6488):97-104.
9
Course of blood pressure in mild hypertensives after withdrawal of long term
antihypertensive treatment. Medical Research Council Working Party on Mild Hypertension.
BMJ. 1986; 293(6553):988-992.
10
Effects of replacing sodium intake in subjects on a low sodium diet: a crossover study.
Australian National Health & Medical Research Council Dietary Salt Study Management
Committee. Clinical and Experimental Hypertension Part A: Theory and Practice. 1989;
11(5-6):1011-1024.
11
Fall in blood pressure with modest reduction in dietary salt intake in mild
hypertension. Australian National Health and Medical Research Council Dietary Salt Study
Management Committee. Lancet. 1989; 1(8635):399-402.
12
Effect of enalapril on survival in patients with reduced left ventricular ejection
fractions and congestive heart failure. The SOLVD Investigators. New England Journal of
Medicine. 1991; 325(5):293-302.
13
Prevention of stroke by antihypertensive drug treatment in older persons with isolated
systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program
(SHEP). SHEP Cooperative Research Group. JAMA. 1991; 265(24):3255-3264.
14
Acute Pain Management: Operative or Medical Procedures and Trauma. Rockville,
MD: Agency for Health Care Policy and Research Publications; 1992
15
Medical Research Council trial of treatment of hypertension in older adults: principal
results. MRC Working Party. BMJ. 1992; 304(6824):405-412.
16
Systolic hypertension in the elderly: Chinese trial (syst-China). Interim report.
Zhonghua Xin Xue Guan Bing Za Zhi. 1992; 20(5):270-5, 323.
17
Systolic hypertension in the elderly: Chinese trial (Syst-China)--second interim report.
Zhonghua Xin Xue Guan Bing Za Zhi. 1993; 21(3):135-7, 185.
18
The Oslo Diet and Exercise Study (ODES): design and objectives. Controlled Clinical
Trials. 1993; 14(3):229-243.
19
Trial of secondary prevention with atenolol after transient ischemic attack or
nondisabling ischemic stroke. The Dutch TIA Trial Study Group. Stroke. 1993; 24(4):543548.
20
Post-stroke antihypertensive treatment study. A preliminary result. PATS
Collaborating Group. Chinese Medical Journal. 1995; 108(9):710-717.
21
BS EN 1060-1: 1996 Specification for Non-Invasive Sphygmomanometers. General
Requirements. 15-6-1996.
22
Clinical and Experimental Pheochromocytoma. Cambridge, MA.: Blackwell; 1996
259
Hypertension (partial update)
Reference list
23
Prevention of coronary heart disease in clinical practice. Recommendations of the
Second Joint Task Force of European and other Societies on coronary prevention. European
Heart Journal. 1998; 19(10):1434-1503.
24
Health Survey for England: Cardiovascular Disease. The Stationary Office; 1999
25
Effects of ramipril on cardiovascular and microvascular outcomes in people with
diabetes mellitus: results of the HOPE study and MICRO-HOPE substudy. Heart Outcomes
Prevention Evaluation Study Investigators. Lancet. 2000; 355(9200):253-259.
26
British National Formulary, Issue 44, March 2003. BMJ Books and Pharmaceutical
Press; 2001
27
Mulrow CD (editors). Evidence Based Hypertension. London: BMJ; 2001
28
Craig R, Mindell J (editors). Health Survey for England 2006. Volume 1:
Cardiovascular Disease and Risk Factors in Adults. Leeds: The Information Centre; 2008.
Available from:
http://www.ic.nhs.uk/webfiles/publications/HSE06/HSE%2006%20report%20VOL%201%20
v2.pdf
29
Principal results of the Japanese trial to assess optimal systolic blood pressure in
elderly hypertensive patients (JATOS). Hypertension Research. 2008; 31(12):2115-2127.
30
NHS Supply Chain catalogue. Available from:
http://www.supplychain.nhs.uk/portal/page/portal/Public/NHS-CAT Last accessed on: 17
February 2010.
31
Achmon J, Granek M, Golomb M, Hart J. Behavioral treatment of essential
hypertension: a comparison between cognitive therapy and biofeedback of heart rate.
Psychosomatic Medicine. 1989; 51(2):152-164.
32
Adsett CA, Bellissimo A, Mitchell A, Wilczynski N, Haynes RB. Behavioral and
physiological effects of a beta blocker and relaxation therapy on mild hypertensives.
Psychosomatic Medicine. 1989; 51(5):523-536.
33
Agras WS, Southam MA, Taylor CB. Long-term persistence of relaxation-induced
blood pressure lowering during the working day. Journal of Consulting and Clinical
Psychology. 1983; 51(5):792-794.
34
Agras WS, Taylor CB, Kraemer HC, Southam MA, Schneider JA. Relaxation training
for essential hypertension at the worksite: II. The poorly controlled hypertensive.
Psychosomatic Medicine. 1987; 49(3):264-273.
35
Aitchison F, Page A. Diagnostic imaging of renal artery stenosis. Journal of Human
Hypertension. 1999; 13(9):595-603.
36
Aivazyan TA, Zaitsev VP, Salenko BB, Yurenev AP, Patrusheva IF. Efficacy of
relaxation techniques in hypertensive patients. Health Psychology. 1988; 7 Suppl:193-200.
37
Alam S, Johnson AG. A meta-analysis of randomised controlled trials (RCT) among
healthy normotensive and essential hypertensive elderly patients to determine the effect of
high salt (NaCl) diet of blood pressure. Journal of Human Hypertension. 1999; 13(6):367374.
38
Alderman MH, Davis TK, Gerber LM, Robb M. Antihypertensive drug therapy
withdrawal in a general population. Archives of Internal Medicine. 1986; 146(7):1309-1311.
39
Alem M, Milia P, Muir S, Lees K, Walters M. Comparison of the effects of diuretics
on blood pressure and arterial stiffness in patients with stroke. Journal of Stroke and
Cerebrovascular Diseases. 2008; 17(6):373-377.
40
American Heart Association. 2000 Heart and Stroke Statistical Update. American
Heart Association, 1999.
41
American National Standard for Electronic or Automated Sphymomanometers.
ANSI/AAMI SP10 1987. Arlington, VA: Association for the Advancement of Medical
Instrumentation, 1983.
260
Hypertension (partial update)
Reference list
42
Amery A, Birkenhager W, Brixko P, Bulpitt C, Clement D, Deruyttere M, De SA,
Dollery C, Fagard R, Forette F, . Mortality and morbidity results from the European Working
Party on High Blood Pressure in the Elderly trial. Lancet. 1985; 1(8442):1349-1354.
43
Amery A, Birkenhager W, Bulpitt CJ, Clement D, De Leeuw P, Dollery CT. Syst-Eur:
a multicenter trial on the treatment of isolated systolic hypertension in the elderly objectives,
protocol, and organisation. Ageing. 1991; 3:287-302.
44
Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular disease risk
profiles. American Heart Journal. 1991; 121(1 Pt 2):293-298.
45
Anderssen S, Holme I, Urdal P, Hjermann I. Diet and exercise intervention have
favourable effects on blood pressure in mild hypertensives: the Oslo Diet and Exercise Study
(ODES). Blood Pressure. 1995; 4(6):343-349.
46
Antivalle M, Lattuada S, Salvaggio A, Paravicini M, Rindi M, Libretti A. Placebo
effect and adaptation to noninvasive monitoring of BP. Journal of Human Hypertension.
1990; 4(6):633-637.
47
Applegate WB, Miller ST, Elam JT, Cushman WC, el Derwi D., Brewer A, Graney
MJ. Nonpharmacologic intervention to reduce blood pressure in older patients with mild
hypertension. Archives of Internal Medicine. 1992; 152(6):1162-1166.
48
Ara, R and Brennan, A. Economic Evaluation of Sibutramine for the Treatment of
Obesity in Adults Without Other Co-Morbidities in the UK. School of Health and Related
Research (ScHARR), University of Sheffield, 2004.
49
Arima H, Chalmers J, Woodward M, Anderson C, Rodgers A, Davis S, Macmahon S,
Neal B. Lower target blood pressures are safe and effective for the prevention of recurrent
stroke: The PROGRESS trial. Journal of Hypertension. 2006; 24(6):1201-1208.
50
Arima H, Tanizaki Y, Yonemoto K, Doi Y, Ninomiya T, Hata J, Fukuhara M,
Matsumura K, Iida M, Kiyohara Y. Impact of blood pressure levels on different types of
stroke: the Hisayama study. Journal of Hypertension. 2009; 27(12):2437-2443.
51
Armitage P, Berry G, Matthews JNS.Statistical Methods in Medical Research. 4th ed.
Oxford: Blackwell Science; 2002
52
Asagami T, Kushiro T, Inoue J, Kanmatsuse K. Long-term reproducibility and
usefulness of daytime recording of noninvasive 24-hour ambulatory blood pressure
monitoring in borderline hypertension: a two-year follow-up study. Clinical and Experimental
Hypertension (New York). 1996; 18(5):637-657.
53
Asayama K, Ohkubo T, Kikuya M, Obara T, Metoki H, Inoue R, Hara A, Hirose T,
Hoshi H, Hashimoto J, Totsune K, Satoh H, Imai Y. Prediction of stroke by home "morning"
versus "evening" blood pressure values: the Ohasama study. Hypertension. 2006; 48(4):737743.
54
Asayama K, Ohkubo T, Yoshida S, Suzuki K, Metoki H, Harada A, Murakami Y,
Ohashi Y, Ueshima H, Imai Y, Japan Arteriosclerosis Longitudinal Study (JALS) group.
Stroke risk and antihypertensive drug treatment in the general population: the Japan
arteriosclerosis longitudinal study. Journal of Hypertension. 2009; 27(2):357-364.
55
ASCOT Steering Committee. Age-Stratified Analysis of Blood Pressure Responses.
Personal communication: 06
56
Asmar R, Safar M, Queneau P. Evaluation of the placebo effect and reproducibility of
blood pressure measurement in hypertension. American Journal of Hypertension. 2001; 14(6
I):546-552.
57
Assmann G, Cullen P, Evers T, Petzinna D, Schulte H. Importance of arterial pulse
pressure as a predictor of coronary heart disease risk in PROCAM. European Heart Journal.
2005; 26(20):2120-2126.
58
Ayman D, Goldshine AD. Blood pressure determination by patients with essential
hypertension: the difference between clinic and home readings before treatment. American
Journal of the Medical Sciences. 1940; 200(4):465-474.
261
Hypertension (partial update)
Reference list
59
Bamford J, Sandercock P, Dennis M, Warlow C, Jones L, McPherson K, Vessey M,
Fowler G, Molyneux A, Hughes T. A prospective study of acute cerebrovascular disease in
the community: the Oxfordshire Community Stroke Project 1981-86. 1. Methodology,
demography and incident cases of first-ever stroke. Journal of Neurology, Neurosurgery &
Psychiatry. 1988; 51(11):1373-1380.
60
Barengo NC, Hu G, Kastarinen M, Antikainen R, Tuomilehto J. The effects of
awareness, treatment and control of hypertension on future stroke incidence in a communitybased population study in Finland. Journal of Hypertension. 2009; 27(7):1459-1465.
61
Barengo NC, Kastarinen M, Antikainen R, Nissinen A, Tuomilehto J. The effects of
awareness, treatment and control of hypertension on cardiovascular and all-cause mortality in
a community-based population. Journal of Human Hypertension. 2009; 23(12):808-816.
62
Bayo J, Cos FX, Roca C, Dalfo A, Martin-Baranera MM, Albert B. Home blood
pressure self-monitoring: Diagnostic performance in white-coat hypertension. Blood Pressure
Monitoring. 2006; 11(2):47-52.
63
Beckett NS, Peters R, Fletcher AE, Staessen JA, Liu L, Dumitrascu D, Stoyanovsky
V, Antikainen RL, Nikitin Y, Anderson C, Belhani A, Forette F, Rajkumar C, Thijs L, Banya
W, Bulpitt CJ, HYVET Study Group. Treatment of hypertension in patients 80 years of age or
older. New England Journal of Medicine. 2008; 358(18):1887-1898.
64
Beevers G, Lip G, O'Brien E. ABC of hypertension: Blood pressure measurement.
Part II. BMJ. 2001; 322(7293):1043-1047.
65
Beevers G, Lip GY, O'Brien E. ABC of hypertension. Blood pressure measurement.
Part I-sphygmomanometry: factors common to all techniques. BMJ. 2001; 322(7292):981985.
66
Beevers G, Lip GY, O'Brien E. ABC of hypertension: The pathophysiology of
hypertension. BMJ. 2001; 322(7291):912-916.
67
Bejan-Angoulvant T, Saadatian-Elahi M, Wright JM, Schron EB, Lindholm LH,
Fagard R, Staessen JA, Gueyffier F. Treatment of hypertension in patients 80 years and older:
The lower the better? A meta-analysis of randomized controlled trials. Journal of
Hypertension. 2010; 28(7):1366-1372.
68
Benetos A, Thomas F, Bean KE, Guize L. Why cardiovascular mortality is higher in
treated hypertensives versus subjects of the same age, in the general population. Journal of
Hypertension. 2003; 21(9):1635-1640.
69
Bennett P, Wallace L, Carroll D, Smith N. Treating Type A behaviours and mild
hypertension in middle-aged men. Journal of Psychosomatic Research. 1991; 35(2-3):209223.
70
Benson J, Britten N. Patients' decisions about whether or not to take antihypertensive
drugs: qualitative study. BMJ. 2002; 325(7369):873.
71
Benson J, Britten N. Patients' views about taking antihypertensive drugs: questionnaire
study. BMJ. 2003; 326(7402):1314-1315.
72
Berglund A, Andersson OK, Berglund G, Fagerberg B. Antihypertensive effect of diet
compared with drug treatment in obese men with mild hypertension. BMJ. 1989;
299(6697):480-485.
73
Bhatnagar D, Anand IS, Durrington PN, Patel DJ, Wander GS, Mackness MI, Creed
F, Tomenson B, Chandrashekhar Y, Winterbotham M, others. Coronary risk factors in people
from the Indian subcontinent living in west London and their siblings in India. Lancet. 1995;
345(8947):405-409.
74
Bhopal R, Unwin N, White M, Yallop J, Walker L, Alberti KG, Harland J, Patel S,
Ahmad N, Turner C, Watson B, Kaur D, Kulkarni A, Laker M, Tavridou A. Heterogeneity of
coronary heart disease risk factors in Indian, Pakistani, Bangladeshi, and European origin
populations: cross sectional study. BMJ. 1999; 319(7204):215-220.
262
Hypertension (partial update)
Reference list
75
Bing RF, Briggs RS, Burden AC, Russell GI, Swales JD, Thurston H. Reversible
hypertension and hypothyroidism. Clinical Endocrinology. 1980; 13(4):339-342.
76
Bing RF, Russell GI, Swales JD, Thurston H. Indapamide and bendrofluazide: A
comparison in the management of essential hypertension. British Journal of Clinical
Pharmacology. 1981; 12(6):883-886.
77
Bjorklund K, Lind L, Zethelius B, Berglund L, Lithell H. Prognostic significance of
24-h ambulatory blood pressure characteristics for cardiovascular morbidity in a population of
elderly men. Journal of Hypertension. 2004; 22(9):1691-1697.
78
Black HR, Elliott WJ, Grandits G, Grambsch P, Lucente T, White WB, Neaton JD,
Grimm RH, Jr., Hansson L, Lacourciere Y, Muller J, Sleight P, Weber MA, Williams G,
Wittes J, Zanchetti A, Anders RJ, CONVINCE Research Group. Principal results of the
Controlled Onset Verapamil Investigation of Cardiovascular End Points (CONVINCE) trial.
JAMA. 2003; 289(16):2073-2082.
79
Black HR, Elliott WJ, Neaton JD, Grandits G, Grambsch P, Grimm RH, Hansson L,
Lacourciere Y, Muller J, Sleight P, Weber MA, White WB, Williams G, Wittes J, Zanchetti
A, Fakouhi TD. Rationale and design for the controlled ONset Verapamil INvestigation of
Cardiovascular Endpoints (CONVINCE) Trial. Controlled Clinical Trials. 1998; 19:370-390.
80
Blanchard EB, Miller ST, Abel GG, Haynes MR, Wicker R. Evaluation of
biofeedback in the treatment of borderline essential hypertension. Journal of Applied
Behavior Analysis. 1979; 12(1):99-109.
81
Bland JM, Jones DR, Bennett S, Haines AP, MacFarlane AJ. Is the clinical trial
evidence about new drugs statistically adequate? British Journal of Clinical Pharmacology.
1985; 19:155-160.
82
Blood Pressure Lowering Treatment Trialists' Collaboration, Turnbull F, Neal B,
Ninomiya T, Algert C, Arima H, Barzi F, Bulpitt C, Chalmers J, Fagard R, Gleason A,
Heritier S, Li N, Perkovic V, Woodward M, Macmahon S. Effects of different regimens to
lower blood pressure on major cardiovascular events in older and younger adults: metaanalysis of randomised trials. BMJ. 2008; 336(7653):1121-1123.
83
Bloom BS. Daily regimen and compliance with treatment. BMJ. 2001; 323(7314):647.
84
Blumenthal JA, Sherwood A, Gullette EC, Babyak M, Waugh R, Georgiades A,
Craighead LW, Tweedy D, Feinglos M, Appelbaum M, Hayano J, Hinderliter A. Exercise and
weight loss reduce blood pressure in men and women with mild hypertension: effects on
cardiovascular, metabolic, and hemodynamic functioning. Archives of Internal Medicine.
2000; 160(13):1947-1958.
85
Blumenthal JA, Siegel WC, Appelbaum M. Failure of exercise to reduce blood
pressure in patients with mild hypertension. Results of a randomized controlled trial. JAMA.
1991; 266(15):2098-2104.
86
Bobrie G, Chatellier G, Genes N, Clerson P, Vaur L, Vaisse B, Menard J, Mallion JM.
Cardiovascular prognosis of "masked hypertension" detected by blood pressure selfmeasurement in elderly treated hypertensive patients. JAMA. 2004; 291(11):1342-1349.
87
Bobrie G, Genes N, Vaur L, Clerson P, Vaisse B, Mallion JM, Chatellier G. Is
"isolated home" hypertension as opposed to "isolated office" hypertension a sign of greater
cardiovascular risk? Archives of Internal Medicine. 2001; 161(18):2205-2211.
88
Boggia J, Li Y, Thijs L, Hansen TW, Kikuya M, Bjorklund-Bodegard K, Richart T,
Ohkubo T, Kuznetsova T, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Wang J, Sandoya E,
O'Brien E, Staessen JA, International Database on Ambulatory blood pressure monitoring in
relation to Cardiovascular Outcomes (IDACO) investigators. Prognostic accuracy of day
versus night ambulatory blood pressure: a cohort study. Lancet. 2007; 370(9594):1219-1229.
89
Borghi C, Dormi A, L'Italien G, Lapuerta P, Franklin SS, Collatina S, Gaddi A. The
relationship between systolic blood pressure and cardiovascular risk--results of the Brisighella
Heart Study. Journal of Clinical Hypertension. 2003; 5(1):47-52.
263
Hypertension (partial update)
Reference list
90
Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa-Terris M, Carr AA,
Kappagoda MD, Rocco MV, Schnaper HW, Sowers JR, Bond MG. Final outcome results of
the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS). JAMA. 1996;
276(10):785-791.
91
Borrello G, Mastroroberto P, Curcio F, Chello M, Zofrea S, Mazza ML. The effects of
magnesium oxide on mild essential hypertension and quality of life. Current Therapeutic
Research. 1996; 57(10):767-774.
92
Bosley F, Allen TW. Stress management training for hypertensives: cognitive and
physiological effects. Journal of Behavioral Medicine. 1989; 12(1):77-89.
93
Bowlus WE, Langford HG. A comparison of the antihypertensive effect of
chlorthalidone and hydrochlorthiazide. Clinical Pharmacology and Therapeutics. 1964;
5:708-711.
94
Brandao SA, Izar CO, Fischer SM, Santos AO, Monteiro CM, Povoa RM, Helfenstein
T, Carvalho AC, Monteiro AM, Ramos E, Gidlund M, Figueiredo Neto AM, Fonseca FA.
Early increase in autoantibodies against human oxidized low-density lipoprotein in
hypertensive patients after blood pressure control. American Journal of Hypertension. 2010;
23(2):208-214.
95
Brauer AP, Horlick L, Nelson E, Farquhar JW, Agras WS. Relaxation therapy for
essential hypertension: a Veterans Administration Outpatient study. Journal of Behavioral
Medicine. 1979; 2(1):21-29.
96
Bray EP, Holder R, Mant J, McManus RJ. Does self-monitoring reduce blood
pressure? Meta-analysis with meta-regression of randomized controlled trials. Annals of
Medicine. 2010; 42(5):371-386.
97
Brenner BM, Cooper ME, de ZD, Keane WF, Mitch WE, Parving HH, Remuzzi G,
Snapinn SM, Zhang Z, Shahinfar S. Effects of losartan on renal and cardiovascular outcomes
in patients with type 2 diabetes and nephropathy. New England Journal of Medicine. 2001;
345(12):861-869.
98
Brenner H, Gefeller O. Variation of sensitivity, specificity, likelihood ratios and
predictive values with disease prevalence. Stat Med. 1997; 16(9):981-991.
99
British Hypertension Society.Blood Pressure Measurement CD-ROM. BMJ; 1999
100 British Hypertension Society. Validated devices suitable for home use. Available
from: http://www.bhsoc.org/blood_pressure_list.stm Last accessed on: 11 November 2010.
101 Britton KA, Gaziano JM, Djousse L. Normal systolic blood pressure and risk of heart
failure in US male physicians. European Journal of Heart Failure. 2009; 11(12):1129-1134.
102 Bronnum-Hansen H, Davidsen M, Thorvaldsen P. Long-term survival and causes of
death after stroke. Stroke. 2001; 32(9):2131-2136.
103 Bronnum-Hansen H, Jorgensen T, Davidsen M, Madsen M, Osler M, Gerdes LU,
Schroll M. Survival and cause of death after myocardial infarction: the Danish MONICA
study. Journal of Clinical Epidemiology. 2001; 54(12):1244-1250.
104 Brook RD. Home blood pressure: accuracy is independent of monitoring schedules.
American Journal of Hypertension. 2000; 13(6 Pt 1):625-631.
105 Brown MJ, Palmer CR, Castaigne A, de Leeuw PW, Mancia G, Rosenthal T.
Morbidity and mortality in patients randomised to double-blind treatment with a long-acting
calcium-channel blocker or diuretic in the International Nifedipine GITS study: Intervention
as a Goal in Hypertension Treatment (INSIGHT). Lancet. 2000; 356(5):366-372.
106 Brown MJ, Palmer CR, Castaigne A, de Leeuw PW, Mancia G, Rosenthal T. Principal
results from the International Nifedipine GITS Study: Intervention as a Goal in Hypertension
Treatment (INSIGHT). European Heart Journal. 2001; 3(Suppl B):B20-B26.
107 Bulpitt CJ, Ferrier G, Lewis PJ, Daymond M, Bulpitt PF, Dollery CT. Potassium
supplementation fails to lower blood pressure in hypertensive patients receiving a potassium
losing diuretic. Annals of Clinical Research. 1985; 17(4):126-130.
264
Hypertension (partial update)
Reference list
108 Burgess E, Lewanczuk R, Bolli P, Chockalingam A, Cutler H, Taylor G, Hamet P.
Lifestyle modifications to prevent and control hypertension. 6. Recommendations on
potassium, magnesium and calcium. Canadian Hypertension Society, Canadian Coalition for
High Blood Pressure Prevention and Control, Laboratory Centre for Disease Control at Health
Canada, Heart and Stroke Foundation of Canada. Canadian Medical Association Journal.
1999; 160(9 Suppl):S35-S45.
109 Burke LE, Dunbar-Jacob JM, Hill MN. Compliance with cardiovascular disease
prevention strategies: a review of the research. Annals of Behavioral Medicine. 1997;
19(3):239-263.
110 Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, Horan MJ,
Labarthe D. Prevalence of hypertension in the US adult population. Results from the Third
National Health and Nutrition Examination Survey, 1988-1991. Hypertension. 1995;
25(3):305-313.
111 Calvo C, Hermida RC, Ayala DE, Lopez JE, Fernandez JR, Dominguez MJ, Mojon A,
Covelo M. The 'ABPM effect' gradually decreases but does not disappear in successive
sessions of ambulatory monitoring. Journal of Hypertension. 2003; 21(12):2265-2273.
112 Campbell NC, Ritchie LD, Thain J, Deans HG, Rawles JM, Squair JL. Secondary
prevention in coronary heart disease: a randomised trial of nurse led clinics in primary care.
Heart. 1998; 80(5):447-452.
113 Campbell NR, Burgess E, Choi BC, Taylor G, Wilson E, Cleroux J, Fodor JG, Leiter
LA, Spence D. Lifestyle modifications to prevent and control hypertension. 1. Methods and
an overview of the Canadian recommendations. Canadian Hypertension Society, Canadian
Coalition for High Blood Pressure Prevention and Control, Laboratory Centre for Disease
Control at Health Canada, Heart and Stroke Foundation of Canada. Canadian Medical
Association Journal. 1999; 160(9 Suppl):S1-S6.
114 Campbell P, Ghuman N, Wakefield D, Wolfson L, White WB. Long-term
reproducibility of ambulatory blood pressure is superior to office blood pressure in the very
elderly. Journal of Human Hypertension. 2010; 24(11):749-754.
115 Canino E, Cardona R, Monsalve P, Perez AF, Lopez B, Fragachan F. A behavioral
treatment program as a therapy in the control of primary hypertension. Acta Cientifica
Venezolana. 1994; 45(1):23-30.
116 Cappuccio FP, Cook DG, Atkinson RW, Strazzullo P. Prevalence, detection, and
management of cardiovascular risk factors in different ethnic groups in south London. Heart.
1997; 78(6):555-563.
117 Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood
pressure? A meta-analysis of published trials. Journal of Hypertension. 1991; 9(5):465-473.
118 Cappuccio FP, Siani A, Strazzullo P. Oral calcium supplementation and blood
pressure: an overview of randomized controlled trials. Journal of Hypertension. 1989;
7(12):941-946.
119 Carlsson AC, Theobald H, Hellénius M, Wändell PE. Cardiovascular and total
mortality in men and women with different blood pressure levels -A 26year followup. Blood
Pressure. 2009; 18(3):105-110.
120 Carson MA, Hathaway A, Tuohey JP, McKay BM. The effect of a relaxation
technique on coronary risk factors. Behavioral Medicine. 1988; 14(2):71-77.
121 Cavallini MC, Roman MJ, Blank SG, Pini R, Pickering TG, Devereux RB.
Association of the auscultatory gap with vascular disease in hypertensive patients. Annals of
Internal Medicine. 1996; 124(10):877-883.
122 Celentano A, Galderisi M, Tammaro P, Mureddu GF, Garofalo M, de Divitiis O.
Effects on casual and 24-h ambulatory blood pressure of slow-release nicardipine and
chlorthalidone in arterial essential hypertension: double-blind, crossover study. International
Journal of Clinical Pharmacology, Therapy, and Toxicology. 1990; 28(5):190-196.
265
Hypertension (partial update)
Reference list
123 Celis H, De CP, Fagard R, Thijs L, Staessen JA. For how many days should blood
pressure be measured at home in older patients before steady levels are obtained? Journal of
Human Hypertension. 1997; 11(10):673-677.
124 Celis H, Yodfat Y, Thijs L, Clement D, Cozic J, de Cort P, Forette F, Grégoire M,
Heyrman J, Stibbe G, van den Haute M, Staessen J, Fagard R, The Syst-Eur Investigators.
Antihypertensive therapy in older patients with isolated systolic hypertension: the Syst-Eur
experience in general practice. Family Practice. 1996; 13(2):138-143.
125 Chalmers J, Morgan T, Doyle A, Dickson B, Hopper J, Mathews J, Matthews G,
Moulds R, Myers J, Nowson C, others. Australian National Health and Medical Research
Council dietary salt study in mild hypertension. Journal of Hypertension - Supplement. 1986;
4(6):S629-S637.
126 Chapman N, Dobson J, Wilson S, Dahlof B, Sever PS, Wedel H, Poulter NR, AngloScandinavian Cardiac Outcomes Trial Investigators. Effect of spironolactone on blood
pressure in subjects with resistant hypertension. Hypertension. 2007; 49(4):839-845.
127 Chaturvedi N, McKeigue PM, Marmot MG. Resting and ambulatory blood pressure
differences in Afro-Caribbeans and Europeans. Hypertension. 1993; 22(1):90-96.
128 Chesney MA, Black GW, Swan GE, Ward MM. Relaxation training for essential
hypertension at the worksite: I. The untreated mild hypertensive. Psychosomatic Medicine.
1987; 49(3):250-263.
129 Chien KL, Huang PJ, Chen MF, Chiang FT, Lai LP, Lee YT. Assessment of quality of
life in a double-blind, randomized clinical trial of imidapril and captopril for hypertensive
Chinese in Taiwan. Cardiovascular Drugs and Therapy. 2002; 16(3):221-226.
130 Chou T. Wake up and smell the coffee. Caffeine, coffee, and the medical
consequences. Western Journal of Medicine. 1992; 157(5):544-553.
131 Clement DL, De Buyzere ML, De Bacquer DA, de Leeuw PW, Duprez DA, Fagard
RH, Gheeraert PJ, Missault LH, Braun JJ, Six RO, Van Der Niepen P, O'Brien E, Office
versus Ambulatory Pressure Study Investigators. Prognostic value of ambulatory bloodpressure recordings in patients with treated hypertension. New England Journal of Medicine.
2003; 348(24):2407-2415.
132 Cleroux J, Feldman RD, Petrella RJ. Lifestyle modifications to prevent and control
hypertension. 4. Recommendations on physical exercise training. Canadian Hypertension
Society, Canadian Coalition for High Blood Pressure Prevention and Control, Laboratory
Centre for Disease Control at Health Canada, Heart and Stroke Foundation of Canada.
Canadian Medical Association Journal. 1999; 160(9 Suppl):S21-S28.
133 Coats AJS, Radaelli A, Clark SJ, Conway J, Sleight P. The influence of ambulatory
blood pressure monitoring on the design and interpretation of trials in hypertension. Journal
of Hypertension. 1992; 10(4):385-391.
134 Coca A, Messerli FH, Benetos A, Zhou Q, Champion A, Cooper-DeHoff RM, Pepine
CJ. Predicting stroke risk in hypertensive patients with coronary artery disease: a report from
the INVEST. Stroke. 2008; 39(2):343-348.
135 Collins R, Peto R, Macmahon S, Hebert P, Fiebach NH, Eberlein KA, Godwin J,
Qizilbash N, Taylor JO, Hennekens CH. Blood pressure, stroke, and coronary heart disease.
Part 2, Short-term reductions in blood pressure: overview of randomised drug trials in their
epidemiological context. Lancet. 1990; 335(8693):827-838.
136 Conen D, Ridker PM, Buring JE, Glynn RJ. Risk of cardiovascular events among
women with high normal blood pressure or blood pressure progression: prospective cohort
study. BMJ. 2007; 335(7617):432.
137 Conen D, Tschudi P, Martina B. Twenty-four hour ambulatory blood pressure for the
management of antihypertensive treatment: a randomized controlled trial. Journal of Human
Hypertension. 2009; 23(2):122-129.
266
Hypertension (partial update)
Reference list
138 Conlin PR, Chow D, Miller ER, III, Svetkey LP, Lin PH, Harsha DW, Moore TJ,
Sacks FM, Appel LJ. The effect of dietary patterns on blood pressure control in hypertensive
patients: results from the Dietary Approaches to Stop Hypertension (DASH) trial. American
Journal of Hypertension. 2000; 13(9):949-955.
139 Cook TD, Campbell DT.Quasi-Experimentation: Design and Analysis Issues for Field
Settings. Chicago: Rand McNally; 1979
140 Coope J, Warrender TS. Randomised trial of treatment of hypertension in elderly
patients in primary care. BMJ. 1986; 293(6555):1145-1151.
141 Costa FV, Ambrosioni E, Montebugnoli L, Paccaloni L, Vasconi L, Magnani B.
Effects of a low-salt diet and of acute salt loading on blood pressure and intralymphocytic
sodium concentration in young subjects with borderline hypertension. Clinical Science. 1981;
61 Suppl 7:21s-23s.
142 Cottier C, Shapiro K, Julius S. Treatment of mild hypertension with progressive
muscle relaxation. Predictive value of indexes of sympathetic tone. Archives of Internal
Medicine. 1984; 144(10):1954-1958.
143 Cramer JA. Consequences of intermittent treatment for hypertension: the case for
medication compliance and persistence. American Journal of Managed Care. 1998;
4(11):1563-1568.
144 Croft PR, Brigg D, Smith S, Harrison CB, Branthwaite A, Collins MF. How useful is
weight reduction in the management of hypertension? Journal of the Royal College of
General Practitioners. 1986; 36(291):445-448.
145 Cruickshank JK, Beevers DG, Osbourne VL, Haynes RA, Corlett JC, Selby S. Heart
attack, stroke, diabetes, and hypertension in West Indians, Asians, and whites in Birmingham,
England. BMJ. 1980; 281(6248):1108.
146 Cupples ME, McKnight A. Randomised controlled trial of health promotion in general
practice for patients at high cardiovascular risk. BMJ. 1994; 309(6960):993-996.
147 Curtin F, Altman DG, Elbourne D. Meta-analysis combining parallel and cross-over
clinical trials. Statistics in Medicine. 2002; 21(15):2131-2144.
148 Cushman WC, Cutler JA, Hanna E, Bingham SF, Follmann D, Harford T, Dubbert P,
Allender PS, Dufour M, Collins JF, Walsh SM, Kirk GF, Burg M, Felicetta JV, Hamilton BP,
Katz LA, Perry HM, Jr., Willenbring ML, Lakshman R, Hamburger RJ. Prevention and
Treatment of Hypertension Study (PATHS): effects of an alcohol treatment program on blood
pressure. Archives of Internal Medicine. 1998; 158(11):1197-1207.
149 Cushman WC, Langford HG. Randomised controlled trial of potassium chloride
versus placebo in mildly hypertensive blacks and whites. Circulation. 1988; 78:Suppl II-370.
150 Cuspidi C, Macca G, Michev I, Salerno M, Fusi V, Severgnini B, Corti C, Meani S,
Valerio C, Magrini F, Zanchetti A. Short-term reproducibility of nocturnal non-dipping
pattern in recently diagnosed essential hypertensives. Blood Pressure. 2002; 11(2):79-83.
151 Cuspidi C, Meani S, Sala C, Valerio C, Fusi V, Zanchetti A, Mancia G. How reliable
is isolated clinical hypertension defined by a single 24-h ambulatory blood pressure
monitoring? Journal of Hypertension. 2007; 25(2):315-320.
152 Cutler JA, Follmann D, Allender PS. Randomized trials of sodium reduction: an
overview. American Journal of Clinical Nutrition. 1997; 65(2 Suppl):643S-651S.
153 Cutler JA, Follmann D, Elliott P, Suh I. An overview of randomized trials of sodium
reduction and blood pressure. Hypertension. 1991; 17(1 Suppl):I27-I33.
154 Dahlöf B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U, Fyhrquist F,
Ibsen H, Kristiansson K, Lederballe-Pedersen O, Lindholm LH, Nieminen M, Omvik S,
Wedel H. Cardiovascular morbidity and mortality in the losartan intervention for endpoint
reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet. 2002;
359:995-1003.
267
Hypertension (partial update)
Reference list
155 Dahlöf B, Hansson L, Lindholm LH, Schersten B, Wester PO, Ekbom T, Hedner T, de
Faire U. STOP-Hypertension-2: a prospective intervention trial of newer versus older
treatment alternatives in old patients with hypertension. Blood Pressure. 1993; 2:136-141.
156 Dahlof B, Lindholm LH, Hansson L, Schersten B, Ekbom T, Wester PO. Morbidity
and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension).
Lancet. 1991; 338(8778):1281-1285.
157 Dahlof B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, Collins R,
Kjeldsen SE, Kristinsson A, McInnes GT, Mehlsen J, Nieminen M, O'Brien E, Ostergren J,
ASCOT Investigators. Prevention of cardiovascular events with an antihypertensive regimen
of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as
required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm
(ASCOT-BPLA): a multicentre randomised controlled trial. Lancet. 2005; 366(9489):895906.
158 Davy BM, Melby CL, Beske SD, Ho RC, Davrath LR, Davy KP. Oat consumption
does not affect resting casual and ambulatory 24-h arterial blood pressure in men with highnormal blood pressure to stage I hypertension. Journal of Nutrition. 2002; 132(3):394-398.
159 Dawes MG, Coats AJ, Juszczak E. Daytime ambulatory systolic blood pressure is
more effective at predicting mortality than clinic blood pressure. Blood Pressure Monitoring.
2006; 11(3):111-118.
160 de Hoon JN, Vanmolkot FH, van den Ven LL, Van Bortel LM. Quality of life
comparison between bisoprolol and nifedipine retard in hypertension. Cardiovascular Drugs
and Therapy. 1997; 11(3):465-471.
161 de Lame PA, Droussin AM, Thomson M, Verhaest L, Wallace S. The effects of
enalapril on hypertension and quality of life. A large multicenter study in Belgium. Acta
Cardiologica. 1989; 44(4):289-302.
162 De Plaen JF, Detry JM. Hemodynamic effects of physical training in established
arterial hypertension. Acta Cardiologica. 1980; 35(3):179-188.
163 de Souza F., Muxfeldt E, Fiszman R, Salles G. Efficacy of spironolactone therapy in
patients with true resistant hypertension. Hypertension. 2010; 55(1):147-152.
164 Deary AJ, Schumann AL, Murfet H, Haydock SF, Foo RS, Brown MJ. Double-blind,
placebo-controlled crossover comparison of five classes of antihypertensive drugs. Journal of
Hypertension. 2002; 20(4):771-777.
165 Deckers JW, Goedhart DM, Boersma E, Briggs A, Bertrand M, Ferrari R, Remme WJ,
Fox K, Simoons ML. Treatment benefit by perindopril in patients with stable coronary artery
disease at different levels of risk. European Heart Journal. 2006; 27(7):796-801.
166 Degl'innocenti A, Elmfeldt D, Hofman A, Lithell H, Olofsson B, Skoog I,
Trenkwalder P, Zanchetti A, Wiklund I. Health-related quality of life during treatment of
elderly patients with hypertension: results from the Study on COgnition and Prognosis in the
Elderly (SCOPE). Journal of Human Hypertension. 2004; 18(4):239-245.
167 Den Hond E, Celis H, Fagard R, Keary L, Leeman M, O'Brien E, Vandenhoven G,
Staessen JA. Self-measured versus ambulatory blood pressure in the diagnosis of
hypertension. Journal of Hypertension. 2003; 21(4):717-722.
168 Denardo SJ, Gong Y, Nichols WW, Messerli FH, Bavry AA, Cooper-DeHoff RM,
Handberg EM, Champion A, Pepine CJ. Blood pressure and outcomes in very old
hypertensive coronary artery disease patients: an INVEST substudy. American Journal of
Medicine. 2010; 123(8):719-726.
169 Dennis M, Bamford J, Sandercock P, Warlow C. Prognosis of transient ischemic
attacks in the Oxfordshire Community Stroke Project. Stroke. 1990; 21(6):848-853.
170 Dennis MS, Bamford JM, Sandercock PA, Warlow CP. Incidence of transient
ischemic attacks in Oxfordshire, England. Stroke. 1989; 20(3):333-339.
268
Hypertension (partial update)
Reference list
171 Department of Health. Blood Pressure Measurement Devices - Mercury and NonMercury. (MDA DB2000(03)). London: Department of Health, 2000.
172 Department of Health. Medical Devices and Equipment Management: Repair and
Maintenance Provision. (MDA DB2000(02)). London: Department of Health, 2000.
173 Department of Health. Exercise Referral Systems: a National Quality Assessment
Framework. London: 2001.
174 Department of Health. NHS reference costs 2008-09. Available from:
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuida
nce/DH_111591
175 DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials.
1986; 7(3):177-188.
176 Devereux RB, Dahlof B, Gerdts E, Boman K, Nieminen MS, Papademetriou V,
Rokkedal J, Harris KE, Edelman JM, Wachtell K. Regression of hypertensive left ventricular
hypertrophy by losartan compared with atenolol: the Losartan Intervention for Endpoint
Reduction in Hypertension (LIFE) trial. Circulation. 2004; 110(11):1456-1462.
177 Dickerson JE, Hingorani AD, Ashby MJ, Palmer CR, Brown MJ. Optimisation of
antihypertensive treatment by crossover rotation of four major classes. Lancet. 1999;
353(9169):2008-2013.
178 Dolan E, Stanton A, Thijs L, Hinedi K, Atkins N, McClory S, Den Hond E,
McCormack P, Staessen JA, O'Brien E. Superiority of ambulatory over clinic blood pressure
measurement in predicting mortality: the Dublin outcome study. Hypertension. 2005;
46(1):156-161.
179 Dolfgoff S, SCHREK R, BALLARD GP, BAKER LA. Tobacco smoking as an
etiologic factor in disease. 2. Coronary disease and hypertension. Angiology. 1952; 3(4):323334.
180 Doll R, HILL AB. Smoking and carcinoma of the lung; preliminary report. BMJ.
1950; 2(4682):739-748.
181 Doll R, Peto R, Hall E, Wheatley K, Gray R. Mortality in relation to consumption of
alcohol: 13 years' observations on male British doctors. BMJ. 1994; 309(6959):911-918.
182 Dosh SA. The diagnosis of essential and secondary hypertension in adults. Journal of
Family Practice. 2001; 50(8):707-712.
183 Drummond MF, Sculpher M, Torrance G, O'Brien B, Stoddart G.Methods for the
Economic Evaluation of Health Care Programmes. 3rd ed. Oxford: OUP; 2005. Available
from:
http://www.oup.com/us/catalog/general/subject/Medicine/PublicHealth/?view=usa&ci=97801
98529453
184 Duncan JJ, Farr JE, Upton SJ, Hagan RD, Oglesby ME, Blair SN. The effects of
aerobic exercise on plasma catecholamines and blood pressure in patients with mild essential
hypertension. JAMA. 1985; 254(18):2609-2613.
185 Dzau VJ, Herrmann HC. Hormonal control of angiotensinogen production. Life
Sciences. 1982; 30(7-8):577-584.
186 Ebrahim S, Smith GD. Systematic review of randomised controlled trials of multiple
risk factor interventions for preventing coronary heart disease. BMJ. 1997; 314(7095):16661674.
187 Ebrahim S, Smith GD. Lowering blood pressure: a systematic review of sustained
effects of non-pharmacological interventions. Journal of Public Health Medicine. 1998;
20(4):441-448.
188 Eccles M, Mason J. How to develop cost-conscious guidelines. Health Technology
Assessment. 2001; 5(16)
269
Hypertension (partial update)
Reference list
189 Egger M, Jüni P, Bartlett C, Holenstein F, Stern J. How important are comprehensive
literature searches and the assessment of trial quality in systematic reviews? Empirical study.
Health Technology Assessment. 2003; 7(1)
190 Eguchi K, Hoshide S, Hoshide Y, Ishikawa S, Shimada K, Kario K. Reproducibility of
ambulatory blood pressure in treated and untreated hypertensive patients. Journal of
Hypertension. 2010; 28(5):918-924.
191 Eguchi K, Kuruvilla S, Ogedegbe G, Gerin W, Schwartz JE, Pickering TG. What is
the optimal interval between successive home blood pressure readings using an automated
oscillometric device? Journal of Hypertension. 2009; 27(6):1172-1177.
192 Eisenberg DM, Delbanco TL, Berkey CS, Kaptchuk TJ, Kupelnick B, Kuhl J,
Chalmers TC. Cognitive behavioral techniques for hypertension: are they effective? Annals of
Internal Medicine. 1993; 118(12):964-972.
193 Elbourne D, Altman DG, Higgins JPT, Curtin F, Worthington HV, vail A. Metaanalyses involving cross-over trials: methodological issues. International Journal of
Epidemiology. 2002; 31(1):140-149.
194 Elliott WJ, Weber RR, Murphy MB. A double-blind, randomized, placebo-controlled
comparison of the metabolic effects of low-dose hydrochlorothiazide and indapamide.
Journal of Clinical Pharmacology. 1991; 31(8):751-757.
195 Emeriau JP, Knauf H, Pujadas JO, Calvo-Gomez C, Abate G, Leonetti G, Chastang C,
European S, I. A comparison of indapamide SR 1.5 mg with both amlodipine 5 mg and
hydrochlorothiazide 25 mg in elderly hypertensive patients: a randomized double-blind
controlled study. Journal of Hypertension. 2001; 19(2):343-350.
196 Enstrom I, Pennert K. Does it matter whether ambulatory blood pressure is recorded
during a work day or a non-work day? Journal of Hypertension. 1996; 14(5):565-569.
197 Eriksson S, Olofsson B, Wester P. Atenolol in Secondary Prevention after Stroke.
Cerebrovascular Diseases. 1995; 5(1):21-25.
198 Ernst ME, Carter BL, Goerdt CJ, Steffensmeier JJG, Phillips BB, Zimmerman MB,
Bergus GR. Comparative antihypertensive effects of hydrochlorothiazide and chlorthalidone
on ambulatory and office blood pressure. Hypertension. 2006; 47(3):352-358.
199 Ernst ME, Carter BL, Zheng S, Grimm RH, Jr. Meta-analysis of dose-response
characteristics of hydrochlorothiazide and chlorthalidone: effects on systolic blood pressure
and potassium. American Journal of Hypertension. 2010; 23(4):440-446.
200 Ernst ME, Weber CA, Dawson JD, O'Connor MA, Lin W, Carter BL, Bergus GR.
How well does a shortened time interval characterize results of a full ambulatory blood
pressure monitoring session? Journal of Clinical Hypertension. 2008; 10(6):431-435.
201 Espeland MA, Whelton PK, Kostis JB, Bahnson JL, Ettinger WH, Cutler JA, Appel
LJ, Kumanyika S, Farmer D, Elam J, Wilson AC, Applegate WB. Predictors and mediators of
successful long-term withdrawal from antihypertensive medications. TONE Cooperative
Research Group. Trial of Nonpharmacologic Interventions in the Elderly. Archives of Family
Medicine. 1999; 8(3):228-236.
202 Esposti LD, Di Martino M., Saragoni S, Sgreccia A, Capone A, Buda S, Esposti ED.
Pharmacoeconomics of antihypertensive drug treatment: an analysis of how long patients
remain on various antihypertensive therapies. Journal of Clinical Hypertension. 2004; 6:7684.
203 Ewald S, vor dem EJ, Uen S, Neikes F, Vetter H, Mengden T. Relationship between
the frequency of blood pressure self-measurement and blood pressure reduction with
antihypertensive therapy : results of the OLMETEL (OLMEsartan TELemonitoring blood
pressure) study. Clinical Drug Investigation. 2006; 26(8):439-446.
204 Fagard RH. The role of exercise in blood pressure control: supportive evidence.
Journal of Hypertension. 1995; 13(11):1223-1227.
270
Hypertension (partial update)
Reference list
205 Fagard RH. Physical activity in the prevention and treatment of hypertension in the
obese. Medicine and Science in Sports and Exercise. 1999; 31(11 Suppl):S624-S630.
206 Fagard RH, Cornelissen VA. Incidence of cardiovascular events in white-coat, masked
and sustained hypertension versus true normotension: A meta-analysis. Journal of
Hypertension. 2007; 25(11):2193-2198.
207 Fagard RH, Staessen JA. Treatment of isolated systolic hypertension in the elderly: the
Syst-Eur trial. Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Clinical and
Experimental Hypertension. 1999; 21(5&6):491-497.
208 Fagard RH, Staessen JA, Thijs L, Bulpitt CJ, Clement D, de Leeuw PW, Jaaskivi M,
Mancia G, O'Brien E, Palatini P, Tuomilehto J, Webster J, Systolic Hypertension in Europe
Trial Investigators. Relationship between ambulatory blood pressure and follow-up clinic
blood pressure in elderly patients with systolic hypertension. Journal of Hypertension. 2004;
22(1):81-87.
209 Fagard RH, Staessen JA, Thijs L, Celis H, Bulpitt CJ, de Leeuw PW, Leonetti G,
Tuomilehto J, Yodfat Y. On-treatment diastolic blood pressure and prognosis in systolic
hypertension. Archives of Internal Medicine. 2007; 167(17):1884-1891.
210 Fagard RH, Thijs L, Staessen JA, Clement DL, De Buyzere ML, De Bacquer DA.
Prognostic significance of ambulatory blood pressure in hypertensive patients with history of
cardiovascular disease. Blood Pressure Monitoring. 2008; 13(6):325-332.
211 Fagard RH, Van Den Broeke C, de Cort P. Prognostic significance of blood pressure
measured in the office, at home and during ambulatory monitoring in older patients in general
practice. Journal of Human Hypertension. 2005; 19(10):801-807.
212 Fagerberg B, Andersson OK, Persson B, Hedner T. Reactivity to norepinephrine and
effect of sodium on blood pressure during weight loss. Hypertension. 1985; 7(4):586-592.
213 Fang X-H, Zhang X-H, Yang Q-D, Dai X-Y, Su F-Z, Rao M-L, Wu S-P, Du X-L,
Wang W-Z, Li S-C. Subtype hypertension and risk of stroke in middle-aged and older
chinese: A 10-year follow-up study. Stroke. 2006; 37(1):38-43.
214 Feldman RD, Campbell N, Larochelle P, Bolli P, Burgess E, Carruthers G, Floras J,
Haynes B, Honos G, Leenen F, Leiter LA, Logan AG, Myers J, Spence D, Zarnke KB.
Lifestyle modifications to prevent and control hypertension. Canadian Medical Association
Journal. 1999; 161(Supplement):S1-S22.
215 Ferrara LA, Iannuzzi R, Castaldo A, Iannuzzi A, Dello RA, Mancini M. Long-term
magnesium supplementation in essential hypertension. Cardiology. 1992; 81(1):25-33.
216 Finnerty FA, Jr. A double-blind study of chlorthalidone and hydrochlorothiazide in an
outpatient population of moderate hypertensives. Angiology. 1976; 27(12):738-744.
217 Fitzgerald PA. Endocrinology. Current Medical Diagnosis and Treatment, 40th edn.
New York: Lange/McGraw-Hill, 2001: 1088-1160.
218 Fletcher AE, Bulpitt CJ, Chase DM, Collins WC, Furberg CD, Goggin TK, Hewett
AJ, Neiss AM. Quality of life with three antihypertensive treatments. Cilazapril, atenolol,
nifedipine. Hypertension. 1992; 19(6 Pt 1):499-507.
219 Flores L, Recasens M, Gomis R, Esmatjes E. White coat hypertension in type 1
diabetic patients without nephropathy. American Journal of Hypertension. 2000; 13(5 Pt
1):560-563.
220 Fodor JG, Whitmore B, Leenen F, Larochelle P. Lifestyle modifications to prevent
and control hypertension. 5. Recommendations on dietary salt. Canadian Hypertension
Society, Canadian Coalition for High Blood Pressure Prevention and Control, Laboratory
Centre for Disease Control at Health Canada, Heart and Stroke Foundation of Canada.
Canadian Medical Association Journal. 1999; 160(9 Suppl):S29-S34.
221 Frankel BL, Patel DJ, Horwitz D, Friedewald WT, Gaarder KR. Treatment of
hypertension with biofeedback and relaxation techniques. Psychosomatic Medicine. 1978;
40(4):276-293.
271
Hypertension (partial update)
Reference list
222 Franklin SS, Wachtell K, Papademetriou V, Olsen MH, Devereux RB, Fyhrquist F,
Ibsen H, Kjeldsen SE, Dahlof B. Cardiovascular morbidity and mortality in hypertensive
patients with lower versus higher risk: a LIFE substudy. Hypertension. 2005; 46(3):492-499.
223 Fravel MA, Ernst ME, Weber CA, Dawson JD, Carter BL, Bergus GR. Influence of
patient characteristics on success of ambulatory blood pressure monitoring.
Pharmacotherapy. 2008; 28(11):1341-1347.
224 Freemantle N, Urdahl H, Eastaugh J, Hobbs FD. What is the place of beta-blockade in
patients who have experienced a myocardial infarction with preserved left ventricular
function? Evidence and (mis)interpretation. Progress in Cardiovascular Diseases. 2002;
44(4):243-250.
225 Fukunaga H, Ohkubo T, Kobayashi M, Tamaki Y, Kikuya M, Obara T, Nakagawa M,
Hara A, Asayama K, Metoki H, Inoue R, Hashimoto J, K, Imai Y. Cost-effectiveness of the
introduction of home blood pressure measurement in patients with office hypertension.
Journal of Hypertension. 2008; 26(4):685-690.
226 Gaddam K, Pimenta E, Thomas SJ, Cofield SS, Oparil S, Harding SM, Calhoun DA.
Spironolactone reduces severity of obstructive sleep apnoea in patients with resistant
hypertension: a preliminary report. Journal of Human Hypertension. 2010; 24(8):532-537.
227 Galloe AM, Graudal N, Moller J, Bro H, Jorgensen M, Christensen HR. Effect of oral
calcium supplementation on blood pressure in patients with previously untreated
hypertension: a randomised, double-blind, placebo-controlled, crossover study. Journal of
Human Hypertension. 1993; 7(1):43-45.
228 Garcia-Vera MP, Sanz J. How many self-measured blood pressure readings are
needed to estimate hypertensive patients' "true" blood pressure? Journal of Behavioral
Medicine. 1999; 22(1):93-113.
229 Geleijnse JM, Witteman JC, Bak AA, den Breeijen JH, Grobbee DE. Reduction in
blood pressure with a low sodium, high potassium, high magnesium salt in older subjects with
mild to moderate hypertension. BMJ. 1994; 309(6952):436-440.
230 Gibbs CR, Beevers DG, Lip GY. The management of hypertensive disease in black
patients. QJM. 1999; 92(4):187-192.
231 Gillett PA, White AT, Caserta MS. Effect of exercise and/or fitness education on
fitness in older, sedentary, obese women. Journal of Aging and Physical Activity. 1996; 4:4245.
232 Goldstein IB, Shapiro D, Thananopavarn C, Sambhi MP. Comparison of drug and
behavioral treatments of essential hypertension. Health Psychology. 1982; 1(1):7-26.
233 Gong L, Zhang W, Zhu Y, Zhu J, Kong D, Page V, Ghadirian P, LeLorier J, Hamet P.
Shanghai trial of nifedipine in the elderly (STONE). Journal of Hypertension. 1996;
14(10):1237-1245.
234 Goodacre S, Nicholl J, Dixon S, Cross E, Angelini K, Arnold J, Revill S, Locker T,
Capewell SJ, Quinney D, Campbell S, Morris F. Randomised controlled trial and economic
evaluation of a chest pain observation unit compared with routine care. British Medical
Journal. 2004; 328(7434):254.
235 Gordon NF, Scott CB, Levine BD. Comparison of single versus multiple lifestyle
interventions: are the antihypertensive effects of exercise training and diet-induced weight
loss additive? American Journal of Cardiology. 1997; 79(6):763-767.
236 Gorton T. Blood pressure of adults by age and sex, United States. Vital and Health
Statistics Series 11, Data From the National Health Survey. 1964; 11:1-40.
237 Gosse P, Cipriano C, Bemurat L, Mas D, Lemetayer P, N'Tela G, Clementy J.
Prognostic significance of blood pressure measured on rising. Journal of Human
Hypertension. 2001; 15(6):413-417.
272
Hypertension (partial update)
Reference list
238 Graudal NA, Galloe AM, Garred P. Effects of sodium restriction on blood pressure,
renin, aldosterone, catecholamines, cholesterols, and triglyceride: a meta-analysis. JAMA.
1998; 279(17):1383-1391.
239 Griffith LE, Guyatt GH, Cook RJ, Bucher HC, Cook DJ. The influence of dietary and
nondietary calcium supplementation on blood pressure: an updated metaanalysis of
randomized controlled trials. American Journal of Hypertension. 1999; 12(1 Pt 1):84-92.
240 Grimm RH, Jr., Neaton JD, Elmer PJ, Svendsen KH, Levin J, Segal M, Holland L,
Witte LJ, Clearman DR, Kofron P, . The influence of oral potassium chloride on blood
pressure in hypertensive men on a low-sodium diet. New England Journal of Medicine. 1990;
322(9):569-574.
241 Grobbee DE, Hofman A. Does sodium restriction lower blood pressure? BMJ. 1986;
293(6538):27-29.
242 Grobbee DE, Hofman A. Effect of calcium supplementation on diastolic blood
pressure in young people with mild hypertension. Lancet. 1986; 2(8509):703-707.
243 Gudmundsson LS, Johannsson M, Thorgeirsson G, Sigfusson N, Sigvaldason H,
Witteman JCM. Hypertension control as predictor of mortality in treated men and women,
followed for up to 30 years. Cardiovascular Drugs and Therapy. 2005; 19(3):227-235.
244 Gustavsen PH, Hoegholm A, Bang LE, Kristensen KS. White coat hypertension is a
cardiovascular risk factor: A 10-year follow-up study. Journal of Human Hypertension. 2003;
17(12):811-817.
245 Hafner RJ. Psychological treatment of essential hypertension: a controlled comparison
of meditation and meditation plus biofeedback. Biofeedback and Self Regulation. 1982;
7(3):305-316.
246 Hagberg JM, Montain SJ, Martin WH, III, Ehsani AA. Effect of exercise training in
60- to 69-year-old persons with essential hypertension. American Journal of Cardiology.
1989; 64(5):348-353.
247 Haider AW, Larson MG, Franklin SS, Levy D. Systolic blood pressure, diastolic
blood pressure, and pulse pressure as predictors of risk for congestive heart failure in the
Framingham Heart Study. Annals of Internal Medicine. 2003; 138(1):10-16.
248 Halbert JA, Silagy CA, Finucane P, Withers RT, Hamdorf PA, Andrews GR. The
effectiveness of exercise training in lowering blood pressure: a meta-analysis of randomised
controlled trials of 4 weeks or longer. Journal of Human Hypertension. 1997; 11(10):641649.
249 Halbert JA, Silagy CA, Finucane PM, Withers RT, Hamdorf PA. Physical activity and
cardiovascular risk factors: effect of advice from an exercise specialist in Australian general
practice. Medical Journal of Australia. 2000; 173(2):84-87.
250 Hall WD. Resistant hypertension, secondary hypertension, and hypertensive crises.
Cardiology Clinics. 2002; 20(2):281-289.
251 Hamet P. The evaluation of the scientific evidence for a relationship between calcium
and hypertension. Journal of Nutrition. 1995; 125(2 Suppl):311S-400S.
252 Hammond IW, Devereux RB, Alderman MH, Lutas EM, Spitzer MC, Crowley JS,
Laragh JH. The prevalence and correlates of echocardiographic left ventricular hypertrophy
among employed patients with uncomplicated hypertension. Journal of the American College
of Cardiology. 1986; 7(3):639-650.
253 Hansen TW, Jeppesen J, Rasmussen S, Ibsen H, Torp-Pedersen C. Ambulatory blood
pressure and mortality: a population-based study. Hypertension. 2005; 45(4):499-504.
254 Hansen TW, Kikuya M, Thijs L, Bjorklund-Bodegard K, Kuznetsova T, Ohkubo T,
Richart T, Torp-Pedersen C, Lind L, Jeppesen J, Ibsen H, Imai Y, Staessen JA, IDACO I.
Prognostic superiority of daytime ambulatory over conventional blood pressure in four
populations: a meta-analysis of 7,030 individuals. Journal of Hypertension. 2007; 25(8):15541564.
273
Hypertension (partial update)
Reference list
255 Hansson L. Results of the STOP-Hypertension-2 trial. Blood Pressure. 2000; 9(Suppl
2):17-20.
256 Hansson L, Hedner T, Lindholm L, et al. The Captopril Prevention Project (CAPPP)
in Hypertension: baseline data and current status. Blood Pressure. 1997; 6:365-367.
257 Hansson L, Hedner T, Lund-Johansen P, Kjeldsen SE, Lindholm LH, Syvertsen JO,
Lanke J, de Faire U, Dahlöf B, Karlberg BE. Randomised trial of effects of calcium
antagonists compared with diuretics and beta-blockers on cardiovascular morbidity and
mortality in hypertension: the Nordic Diltiazem (NORDIL) study. Lancet. 2000; 356:359365.
258 Hansson L, Lindholm LH, Ekbom T, Dahlöf B, Lanke J, Schersten B, Wester PO,
Hedner T. Randomised trial of old and new antihypertensive drugs in elderly patients:
cardiovascular mortality and morbidity the Swedish Trial in Old Patients with Hypertension-2
study. Lancet. 1999; 354:1751-1756.
259 Hansson L, Lindholm LH, Niskanen L, Lanke J, Hedner T, Niklason A, Luomanmaki
K, Dahlof B, de FU, Morlin C, Karlberg BE, Wester PO, Bjorck JE. Effect of angiotensinconverting-enzyme inhibition compared with conventional therapy on cardiovascular
morbidity and mortality in hypertension: the Captopril Prevention Project (CAPPP)
randomised trial. Lancet. 1999; 353(9153):611-616.
260 Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S, Menard J,
Rahn KH, Wedel H, Westerling S. Effects of intensive blood-pressure lowering and low-dose
aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment
(HOT) randomised trial. HOT Study Group. Lancet. 1998; 351(9118):1755-1762.
261 Harris KA, Holly RG. Physiological response to circuit weight training in borderline
hypertensive subjects. Medicine and Science in Sports and Exercise. 1987; 19(3):246-252.
262 Harsha DW, Lin PH, Obarzanek E, Karanja NM, Moore TJ, Caballero B. Dietary
Approaches to Stop Hypertension: a summary of study results. DASH Collaborative Research
Group. Journal of the American Dietetic Association. 1999; 99(8 Suppl):S35-S39.
263 Hart CL, Smith GD, Hole DJ, Hawthorne VM. Alcohol consumption and mortality
from all causes, coronary heart disease, and stroke: results from a prospective cohort study of
scottish men with 21 years of follow up. BMJ. 1999; 318(7200):1725-1729.
264 Harvard CEA Registry. Cost Effectiveness Analysis (CEA) Registry. 1997. TuftsNew England Medical Center. https://research.tufts-nemc.org/cear4/Default.aspx
265 Hatch JP, Klatt KD, Supik JD, Rios N, Fisher JG, Bauer RL, Shimotsu GW.
Combined behavioral and pharmacological treatment of essential hypertension. Biofeedback
and Self Regulation. 1985; 10(2):119-138.
266 Hatt PY, Leblond JB. A comparative study of the activity of a new agent, indapamide,
in essential arterial hypertension. Current Medical Research and Opinion. 1975; 3(3):138144.
267 Haynes RB, McDonald H, Garg AX, Montague P. Interventions for helping patients to
follow prescriptions for medications. Cochrane Database of Systematic Reviews.
2002;(2):CD000011.
268 He FJ, MacGregor GA. Effect of modest salt reduction on blood pressure: a metaanalysis of randomized trials. Implications for public health. Journal of Human Hypertension.
2002; 16(11):761-770.
269 Head GA, Mihailidou AS, Duggan KA, Beilin LJ, Berry N, Brown MA, Bune AJ,
Cowley D, Chalmers JP, Howe PRC, Hodgson J, Ludbrook J, Mangoni AA, McGrath BP,
Nelson MR, Sharman JE, Stowasser M. Definition of ambulatory blood pressure targets for
diagnosis and treatment of hypertension in relation to clinic blood pressure: Prospective
cohort study. BMJ. 2010; 340(7751):849.
274
Hypertension (partial update)
Reference list
270 Henderson DG, Schierup J, Schodt T. Effect of magnesium supplementation on blood
pressure and electrolyte concentrations in hypertensive patients receiving long term diuretic
treatment. BMJ. 1986; 293(6548):664-665.
271 Hermida RC, Calvo C, Ayala DE, Fernandez JR, Ruilope LM, Lopez JE. Evaluation
of the extent and duration of the "ABPM effect" in hypertensive patients. Journal of the
American College of Cardiology. 2002; 40(4):710-717.
272 Hernandez-del RR, Martin-Baranera M, Sobrino J, Gorostidi M, Vinyoles E, Sierra C,
Segura J, Coca A, Ruilope LM, Spanish Society of Hypertension Ambulatory Blood Pressure
Monitoring Registry Investigators. Reproducibility of the circadian blood pressure pattern in
24-h versus 48-h recordings: the Spanish Ambulatory Blood Pressure Monitoring Registry.
Journal of Hypertension. 2007; 25(12):2406-2412.
273 Hershey JC, Morton BG, Davis JB, Reichgott MJ. Patient compliance with
antihypertensive medication. American Journal of Public Health. 1980; 70(10):1081-1089.
274 Hinderliter A, Sherwood A, Gullette EC, Babyak M, Waugh R, Georgiades A,
Blumenthal JA. Reduction of left ventricular hypertrophy after exercise and weight loss in
overweight patients with mild hypertension. Archives of Internal Medicine. 2002;
162(12):1333-1339.
275 Hodgkinson J, Mant J, Martin U, Guo B, Hobbs FDR, Deeks JJ, Heneghan C, Roberts
N, McManus RJ. Relative effectiveness of clinic and home blood pressure monitoring
compared with ambulatory blood pressure monitoring in diagnosis of hypertension:
systematic review. BMJ. 2011; 342:d3621
276 Hoelscher TJ, Lichstein KL, Fischer SM, Hegarty TB. Relaxation treatment of
hypertension: Do home relaxation tapes enhance treatment outcome? Behavior Therapy.
1987; 18(1):33-37.
277 Hoelscher TJ, Lichstein KL, Rosenthal TL. Home relaxation practice in hypertension
treatment: objective assessment and compliance induction. Journal of Consulting and Clinical
Psychology. 1986; 54(2):217-221.
278 Hollenberg NK, Williams GH, Anderson R, Akhras KS, Bittman RM, Krause SL.
Symptoms and the distress they cause: comparison of an aldosterone antagonist and a calcium
channel blocking agent in patients with systolic hypertension. Archives of Internal Medicine.
2003; 163(13):1543-1548.
279 Hooper L, Bartlett C, Davey SG, Ebrahim S. Systematic review of long term effects of
advice to reduce dietary salt in adults. BMJ. 2002; 325(7365):628.
280 Hosohata K, Saito S, Ohkubo T, Kikuya M, Obara T, Kato T, Totsune K, Miura Y,
Arakawa K, Fujishima M, Fujii J, Fukiyama K, Hisamichi S, Iimura O, Ishii M, Omae T,
Saruta T, Yoshinaga K, Abe I, Abukawa T, Ashida T, Dohba N, Etoh T, Fujimura A, Gotoh
T, Hama H, Hano T, Hayashi H, Hayashida N, Hayashi M, Hiramori K, Hirai Y, Hirata Y,
Hiwada K, Hora K, Ichikawa S, Imaizumi T, Ishikawa K, Ito I, Iwaoka D, Kanamasa K,
Katagiri T, Katayama S, Kawano Y, Kida H, Kimura G, Kitaoka H, Kobayashi S, Kohara K,
Kojima S, Komuro I, Kumagai H, Kusano E, Kushiro T, Kuwajima I, Maruyama Y, Masani
F, Matsubara H, Matsubara T, Matsumoto M, Matsuoka H, Matsuura H, Mishima Y, Miura
M, Miyamori I, Murakami H, Muratani H, Nakao K, Naruse M, Nishio I, Ogihara T, Ohta M,
Ohtsuka K, Ohuchi Y, Oikawa S, Okabe M, Okumura K, Saitoh I, Saitoh H, Sakata T, Saku
K, Sasaki H, Senda S, Shimada K, Shimamoto K, Shiomi T, Shirato K, Takada M, Takeda K,
Takeda N, Takeshita A, Takishita S, Toba K, Tochikubo O, Tomoike H, Ueno Y, Umemura
S, Urata H, Yamada K, Yamaguchi T, Yamashina A, Yoshimura M, Satoh H, Oka Y, Katagiri
H, Kondo Y, Shishido H, Kohinata A, Kanno Y, Ikeda H, Takahashi T, Kimura A, Funahashi
J, Takada N, Hanazawa T, Osugi M, Hashimoto J, Miyamura T, Masuda Y, Tanaka S,
Koizumi M, Ando T, Tsuchida A, Nanba M, Yonekura S, Shiroishi M, Nakagawa M, Oda K,
Kanahara I, Hirota N, Tsutsui M, Anzai J, Hosoda S, Shiiki M, Takagawa Y, Iida M, Tada M,
Noto T, Tanaka H, Tajima G, Kawamorita K, Komai K, Seki H, Omoto A, Takahashi H,
275
Hypertension (partial update)
Reference list
Kitabayashi A, Kimura M, Kimura Y, Takahashi K, Kawamorita Y, Emura Y, Ishikawa Y,
Nagai K, Nagai S, Ito T, Nihei K, Yamada S, Yamamoto N, Suzuki Y, Ito H, Ito M, Fujita T,
Tominaga S, Suzuki K, Ishibashi K, Ando Y, Sato S, Morikawa H, Kashima S, Nakayama D,
Hayashi Y, Ohta K, Metoki H, Fukami K, Hayashi T, Kishi M, Tajima J, Seino J, Hoshi H,
Imai Y, Kimura H, Mori R, Matsuo K, Tanno Y, Shibasaki A, Yamagishi T, Ohtomo E,
Ohtsuka Y, Sasaki S, Seino M, Kurosawa K, Kyogoku S, Ito K, Ono Y, Watanabe S,
Hiwatashi N, Yagi C, Unakami H, Asayama K, Mouri T, Watanabe T, Kikuchi K, Maruyama
N, Sone M, Iwamoto M, Naganuma S, Mashiko H, Ishii H, Takaya Y, Kamimoto M, Shirai
A, Watanabe R, Tohyama Y, Araki F, Sakuma H, Amada Y, Sato E, Techigawara M,
Nakamura K, Nakayama T, Hasegawa K, Minami J, Kumagai Y, Cho T, Okamoto K, Iguchi
T, Honzawa T, Koitabashi T, Horikoshi H, Nagao T, Yoshimatsu H, Seki K, Akiyama N,
Suga H, Matsunaga R, Sinozaki T, Ishimaru Y, Yagi N, Takikawa H, Hukutome T, Nakajima
C, Sato H, Simizu J, Arai M, Koide H, Takada H, Umezu T, Aoki H, Nakamoto H,
Komahashi K, Sekiya S, Sugimoto H, Arai T, Inokuma S, Funazaki T. Progress report on the
Hypertension Objective Treatment Based on Measurement by Electrical Devices of Blood
Pressure (HOMED-BP) study: Status at February 2004. Clinical and Experimental
Hypertension. 2007; 29(1):69-81.
281 Hulley SB, Furberg CD, Gurland B, McDonald R, Perry HM, Schnaper HW. Systolic
Hypertension in the Elderly Program (SHEP): antihypertensive efficacy of chlorthalidone.
American Journal of Cardiology. 1985; 56(15):913-920.
282 Ichihara A, Hayashi M, Koura Y, Tada Y, Hirota N, Saruta T. Long-term effects of
intensive blood-pressure lowering on arterial wall stiffness in hypertensive patients. American
Journal of Hypertension. 2003; 16(11 Pt 1):959-965.
283 Imai Y, Satoh H, Nagai K, Sakuma M, Sakuma H, Minami N, Munakata M,
Hashimoto J, Yamagishi T, Watanabe N, . Characteristics of a community-based distribution
of home blood pressure in Ohasama in northern Japan. Journal of Hypertension. 1993;
11(12):1441-1449.
284 Ingelsson E, Bjorklund-Bodegard K, Lind L, Arnlov J, Sundstrom J. Diurnal blood
pressure pattern and risk of congestive heart failure. JAMA. 2006; 295(24):2859-2866.
285 Inoue R, Ohkubo T, Kikuya M, Metoki H, Asayama K, Obara T, Hirose T, Hara A,
Hoshi H, Hashimoto J, Totsune K, Satoh H, Kondo Y, Imai Y. Stroke risk in systolic and
combined systolic and diastolic hypertension determined using ambulatory blood pressure.
The Ohasama study. American Journal of Hypertension. 2007; 20(10):1125-1131.
286 Institute for Clinical Systems Improvement. Hypertension Diagnosis and Treatment.
Bloomington, MN: Institute for Clinical Systems Improvement, 2000.
287 International Centre for Alcohol Policies. What Is a "Standard Drink"? (5).
Washington, DC: International Centre for Alcohol Policies, 1998.
288 Irvine MJ, Johnston DW, Jenner DA, Marie GV. Relaxation and stress management in
the treatment of essential hypertension. Journal of Psychosomatic Research. 1986; 30(4):437450.
289 Irvine MJ, Logan AG. Relaxation behavior therapy as sole treatment for mild
hypertension. Psychosomatic Medicine. 1991; 53(6):587-597.
290 Ishikawa J, Carroll DJ, Kuruvilla S, Schwartz JE, Pickering TG. Changes in home
versus clinic blood pressure with antihypertensive treatments: a meta-analysis. Hypertension.
2008; 52:856-864.
291 Ishikawa S, Kario K, Kayaba K, Gotoh T, Nago N, Nakamura Y, Tsutsumi A, Kajii E.
Continued high risk of stroke in treated hypertensives in a general population: The Jichi
medical school cohort study. Hypertension Research. 2008; 31(6):1125-1133.
292 Iskedjian M, Einarson TR, MacKeigan LD, Shear N, Addis A, Mittmann N, Ilersich
AL. Relationship between daily dose frequency and adherence to antihypertensive
276
Hypertension (partial update)
Reference list
pharmacotherapy: evidence from a meta-analysis. Clinical Therapeutics. 2002; 24(2):302316.
293 Jacob RG, Chesney MA, Williams DM, Ding Y, Shapiro AP. Relaxation therapy for
hypertension: design effects and treatment effects. Annals of Behavioral Medicine. 1991;
13(1):5-17.
294 Jacob RG, Fortmann SP, Kraemer HC, Farquhar JW, Agras WS. Combining
behavioral treatments to reduce blood pressure. A controlled outcome study. Behavior
Modification. 1985; 9(1):32-53.
295 Jalkanen L. The effect of a weight reduction program on cardiovascular risk factors
among overweight hypertensives in primary health care. Scandinavian Journal of Social
Medicine. 1991; 19(1):66-71.
296 Jamerson K, Weber MA, Bakris GL, Dahlof B, Pitt B, Shi V, Hester A, Gupte J,
Gatlin M, Velazquez EJ, ACCOMPLISH Trial Investigators. Benazepril plus amlodipine or
hydrochlorothiazide for hypertension in high-risk patients. New England Journal of Medicine.
2008; 359(23):2417-2428.
297 James JE. Is habitual caffeine use a preventable cardiovascular risk factor? Lancet.
1997; 349(9047):279-281.
298 JATOS Study Group. The Japanese Trial to Assess Optimal Systolic Blood Pressure
in Elderly Hypertensive Patients (JATOS): protocol, patient characteristics, and blood
pressure during the first 12 months. Hypertension Research. 2005; 28(6):513-520.
299 Jee SH, He J, Whelton PK, Suh I, Klag MJ. The effect of chronic coffee drinking on
blood pressure: a meta-analysis of controlled clinical trials. Hypertension. 1999; 33(2):647652.
300 Jee SH, Miller ER, III, Guallar E, Singh VK, Appel LJ, Klag MJ. The effect of
magnesium supplementation on blood pressure: a meta-analysis of randomized clinical trials.
American Journal of Hypertension. 2002; 15(8):691-696.
301 Johannesson M. The cost-effectiveness of th switch towards more expensive
antihypertensive drugs. Health Policy. 1994; 28(1):1-13.
302 Johansson JK, Niiranen TJ, Puukka PJ, Jula AM. Optimal schedule for home blood
pressure monitoring based on a clinical approach. Journal of Hypertension. 2010; 28(2):259264.
303 Johnson KA, Partsch DJ, Rippole LL, McVey DM. Reliability of self-reported blood
pressure measurements. Archives of Internal Medicine. 1999; 159(22):2689-2693.
304 Johnston DW, Gold A, Kentish J, Smith D, Vallance P, Shah D, Leach G, Robinson
B. Effect of stress management on blood pressure in mild primary hypertension. BMJ. 1993;
306(6883):963-966.
305 Johnston ME, Gibson ES, Terry CW, Haynes RB, Taylor DW, Gafni A, Sicurella JI,
Sackett DL. Effects of labelling on income, work and social function among hypertensive
employees. Journal of Chronic Diseases. 1984; 37(6):417-423.
306 Joint Formulary Committee.British National Formulary: BNF 60. 60th ed. London:
UK: BMJ Group and Pharmaceutical Press; 2010. Available from: http://www.bnf.org.uk
307 Jonsson B, Carides GW, Burke TA, Dasbach EJ, Lindholm LH, Dahlof B. Cost
effectiveness of losartan in patients with hypertension and LVH: an economic evaluation for
Sweden of the LIFE trial. Journal of Hypertension. 2005; 23(7):1425-1431.
308 Jonsson B, Hansson L, Stalhammar NO. Health economics in the hypertension
optimal treatment (HOT) study: costs and cost-effectiveness of intensive blood pressure
lowering and low-dose aspirin in patients with hypertension. Journal of Internal Medicine.
2003; 253:472-480.
309 Jula A, Ronnemaa T, Rastas M, Karvetti RL, Maki J. Long-term nopharmacological
treatment for mild to moderate hypertension. Journal of Internal Medicine. 1990; 227(6):413421.
277
Hypertension (partial update)
Reference list
310 Jula A, Ronnemaa T, Tikkanen I, Karanko H. Responses of atrial natriuretic factor to
long-term sodium restriction in mild to moderate hypertension. Journal of Internal Medicine.
1992; 231(5):521-529.
311 Jula AM, Karanko HM. Effects on left ventricular hypertrophy of long-term
nonpharmacological treatment with sodium restriction in mild-to-moderate essential
hypertension. Circulation. 1994; 89(3):1023-1031.
312 Julius S, Kjeldsen SE, Weber M, Brunner HR, Ekman S, Hansson L, Hua T, Laragh J,
McInnes GT, Mitchell L, Plat F, Schork A, Smith B, Zanchetti A, VALUE trial group.
Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on
valsartan or amlodipine: the VALUE randomised trial. Lancet. 2004; 363(9426):2022-2031.
313 Kagiyama S, Fukuhara M, Ansai T, Matsumura K, Soh I, Takata Y, Sonoki K, Awano
S, Takehara T, Iida M. Association between blood pressure and mortality in 80-year-old
subjects from a population- based prospective study in Japan. Hypertension Research. 2008;
31(2):265-270.
314 Kaplan NM. Other Forms of Secondary Hypertension. In: Kaplan NM, Lieberman E
(eds), Clinical Hypertension, 7th edition edn. Baltimore: Williams & Wilkins, 1998:
315 Kawabe H, Saito I. Correlation of repeated measurements of home blood pressure on
one occasion and diagnosis of hypertension: study by measurement over seven consecutive
days. Clinical and Experimental Hypertension. 2008; 30(1):79-85.
316 Kawabe H, Saito I, Saruta T. Influence of repeated measurement on one occasion, on
successive days, and on workdays on home blood pressure values. Clinical and Experimental
Hypertension. 2005; 27(2-3):215-222.
317 Kawano Y, Matsuoka H, Takishita S, Omae T. Effects of magnesium supplementation
in hypertensive patients: assessment by office, home, and ambulatory blood pressures.
Hypertension. 1998; 32(2):260-265.
318 Kawano Y, Yoshimi H, Matsuoka H, Takishita S, Omae T. Calcium supplementation
in patients with essential hypertension: assessment by office, home and ambulatory blood
pressure. Journal of Hypertension. 1998; 16(11):1693-1699.
319 Kelley GA. Aerobic exercise and resting blood pressure among women: a metaanalysis. Preventive Medicine. 1999; 28(3):264-275.
320 Kelley GA, Kelley KS. Aerobic exercise and resting blood pressure in women: a
meta-analytic review of controlled clinical trials. Journal of Women's Health & Gender-Based
Medicine. 1999; 8(6):787-803.
321 Kelley GA, Kelley KS. Progressive resistance exercise and resting blood pressure : A
meta-analysis of randomized controlled trials. Hypertension. 2000; 35(3):838-843.
322 Kelley GA, Kelley KS, Tran ZV. Walking and resting blood pressure in adults: a
meta-analysis. Preventive Medicine. 2001; 33(2 Pt 1):120-127.
323 Kelley GA, Sharpe KK. Aerobic exercise and resting blood pressure in older adults: a
meta-analytic review of randomized controlled trials. Journals of Gerontology Series A,
Biological Sciences and Medical Sciences. 2001; 56(5):M298-M303.
324 Khattar RS, Swales JD, Banfield A, Dore C, Senior R, Lahiri A. Prediction of
coronary and cerebrovascular morbidity and mortality by direct continuous ambulatory blood
pressure monitoring in essential hypertension. Circulation. 1999; 100(10):1071-1076.
325 Khattar RS, Swales JD, Dore C, Senior R, Lahiri A. Effect of aging on the prognostic
significance of ambulatory systolic, diastolic, and pulse pressure in essential hypertension.
Circulation. 2001; 104(7):783-789.
326 Kikuya M, Hansen TW, Thijs L, Bjorklund-Bodegard K, Kuznetsova T, Ohkubo T,
Richart T, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Staessen JA, International Database on
Ambulatory Blood Pressure Monitoring in relation to Cardiovascular Outcomes Investigators.
Diagnostic thresholds for ambulatory blood pressure monitoring based on 10-year
cardiovascular risk. Circulation. 2007; 115(16):2145-2152.
278
Hypertension (partial update)
Reference list
327 Kinoshita A, Urata H, Tanabe Y, Ikeda M, Tanaka H, Shindo M, Arakawa K. What
types of hypertensives respond better to mild exercise therapy? Journal of Hypertension Supplement. 1988; 6(4):S631-S633.
328 Kjeldsen SE, Lyle PA, Kizer JR, Dahlof B, Devereux RB, Julius S, Beevers G, de
Faire U, Fyhrquist F, Ibsen H, Kristianson K, Lederballe-Pedersen O, Lindholm LH,
Nieminen MS, Omvik P, Oparil S, Snapinn SM, Harris KE, Wedel H, LIFE Study Group. The
effects of losartan compared to atenolol on stroke in patients with isolated systolic
hypertension and left ventricular hypertrophy. The LIFE study. Journal of Clinical
Hypertension. 2005; 7(3):152-158.
329 Klein A. Thyroid Hormone and High Blood Pressure. In: Laragh JH, Brener BM,
Kaplan NM (eds), Endocrine Mechanism in Hypertension, New York: Raven Press, 1989: 6179.
330 Koehler E, Brown E, Haneuse J-PA. On the Assessment of Monte Carlo Error in
Simulation-Based Statistical Analyses. The American Statistician. 2009; 63(2):155-162.
331 Kokubo Y, Kamide K, Okamura T, Watanabe M, Higashiyama A, Kawanishi K,
Okayama A, Kawano Y. Impact of high-normal blood pressure on the risk of cardiovascular
disease in a Japanese urban cohort: the Suita study. Hypertension. 2008; 52(4):652-659.
332 Kono S, Kushiro T, Hirata Y, Hamada C, Takahashi A, Yoshida Y. Class of
antihypertensive drugs, blood pressure status, and risk of cardiovascular disease in
hypertensive patients: A case-control study in Japan. Hypertension Research. 2005;
28(10):811-817.
333 Koopman H, Deville W, van Eijk JT, Donker AJ, Spreeuwenberg C. Diet or diuretic?
Treatment of newly diagnosed mild to moderate hypertension in the elderly. Journal of
Human Hypertension. 1997; 11(12):807-812.
334 Korhonen M, Kastarinen M, Uusitupa M, Puska P, Nissinen A. The effect of
intensified diet counseling on the diet of hypertensive subjects in primary health care: a 2-year
open randomized controlled trial of lifestyle intervention against hypertension in eastern
Finland. Preventive Medicine. 2003; 36(1):8-16.
335 Kostis JB, Davis BR, Cutler J, Grimm RH, Jr., Berge KG, Cohen JD, Lacy CR, Perry
HM, Jr., Blaufox MD, Wassertheil-Smoller S, Black HR, Schron E, Berkson DM, Curb JD,
Smith WM, McDonald R, Applegate WB. Prevention of heart failure by antihypertensive
drug treatment in older persons with isolated systolic hypertension. SHEP Cooperative
Research Group. JAMA. 1997; 278(3):212-216.
336 Kostis JB, Espeland MA, Appel L, Johnson KC, Pierce J, Wofford JL. Does
withdrawal of antihypertensive medication increase the risk of cardiovascular events? Trial of
Nonpharmacologic Interventions in the Elderly (TONE) Cooperative Research Group.
American Journal of Cardiology. 1998; 82(12):1501-1508.
337 Kostis JB, Rosen RC, Brondolo E, Taska L, Smith DE, Wilson AC. Superiority of
nonpharmacologic therapy compared to propranolol and placebo in men with mild
hypertension: a randomized, prospective trial. American Heart Journal. 1992; 123(2):466474.
338 Krakoff LR. Cost-effectiveness of ambulatory blood pressure: a reanalysis.
Hypertension. 2006; 47(1):29-34.
339 Kreeft JH, Langlois S, Ogilvie RI. Comparative trial of indapamide and
hydrochlorothiazide in essential hypertension, with forearm plethysmography. Journal of
Cardiovascular Pharmacology. 1984; 6(4):622-626.
340 Kshirsagar AV, Carpenter M, Bang H, Wyatt SB, Colindres RE. Blood pressure
usually considered normal is associated with an elevated risk of cardiovascular disease.
American Journal of Medicine. 2006; 119(2):133-141.
279
Hypertension (partial update)
Reference list
341 Kukkonen K, Rauramaa R, Voutilainen E, Lansimies E. Physical training of middleaged men with borderline hypertension. Annals of Clinical Research. 1982; 14 Suppl 34:139145.
342 Kumanyika SK, Adams-Campbell L, Van HB, Ten Have TR, Treu JA, Askov E,
Williams J, Achterberg C, Zaghloul S, Monsegu D, Bright M, Stoy DB, Malone-Jackson M,
Mooney D, Deiling S, Caulfield J. Outcomes of a cardiovascular nutrition counseling
program in African-Americans with elevated blood pressure or cholesterol level. Journal of
the American Dietetic Association. 1999; 99(11):1380-1391.
343 Kuwajima I, Kuramoto K, Ogihara T, Iimura O, Abe K, Saruta T, et al. Tolerability
and safety of a calcium channel blocker in comparison with a diuretic in the treatment of
elderly patients with hypertension: Secondary analysis of the NICS-EH. Hypertension
Research. 2001; 24:475-480.
344 Kyngas H, Lahdenpera T. Compliance of patients with hypertension and associated
factors. Journal of Advanced Nursing. 1999; 29(4):832-839.
345 LaGrone R, Jeffrey TB, Ferguson CL. Effects of education and relaxation training
with essential hypertension patients. Journal of Clinical Psychology. 1988; 44(2):271-276.
346 Lahdenpera T, Kyngas H. Compliance and its evaluation in patients with
hypertension. Journal of Clinical Nursing. 2000; 9:826-833.
347 Lane DA, Shah S, Beevers DG. Low-dose spironolactone in the management of
resistant hypertension: A surveillance study. Journal of Hypertension. 2007; 25(4):891-894.
348 Lang T, Nicaud V, Darne B, Rueff B. Improving hypertension control among exssive
alcohol drinkers: a randomised controlled trial in France. The Walpa Group. Journal of
Epidemiology and Community Health. 1995; 49(610):616.
349 Langford HG, Blaufox MD, Oberman A, Hawkins CM, Curb JD, Cutter GR,
Wassertheil-Smoller S, Pressel S, Babcock C, Abernethy JD, . Dietary therapy slows the
return of hypertension after stopping prolonged medication. JAMA. 1985; 253(5):657-664.
350 Law MR, Frost CD, Wald NJ. By how much does dietary salt reduction lower blood
pressure? III--Analysis of data from trials of salt reduction. BMJ. 1991; 302(6780):819-824.
351 Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention
of cardiovascular disease: Meta-analysis of 147 randomised trials in the context of
expectations from prospective epidemiological studies. BMJ. 2009; 338(7705):1245.
352 Law MR, Morris JK and Wald NJ. Unpublished Data From the Published Study: Law
MR, Morris JK, Wald NJ. Use of Blood Pressure Lowering Drugs in the Prevention of
Cardiovascular Disease: Meta-Analysis of 147 Randomised Trials in the Context of
Expectations From Prospective Epidemiological Studies. BMJ. 2009; 338(7705):1245.
Personal communication: 22/11/10
353 Lede RL, Voto LS, Orti J, Margulies M. Agreement between different frequencies of
measurements in ambulatory blood pressure monitoring. Journal of Obstetrics and
Gynaecology. 1997; 17(4):337-339.
354 Leenen FH, Nwachuku CE, Black HR, Cushman WC, Davis BR, Simpson LM,
Alderman MH, Atlas SA, Basile JN, Cuyjet AB, Dart R, Felicetta JV, Grimm RH, Haywood
LJ, Jafri SZ, Proschan MA, Thadani U, Whelton PK, Wright JT, Antihypertensive and LipidLowering Treatment to Prevent Heart Attack Trial Collaborative Research Group. Clinical
events in high-risk hypertensive patients randomly assigned to calcium channel blocker
versus angiotensin-converting enzyme inhibitor in the antihypertensive and lipid-lowering
treatment to prevent heart attack trial. Hypertension. 2006; 48(3):374-384.
355 Leiter LA, Abbott D, Campbell NR, Mendelson R, Ogilvie RI, Chockalingam A.
Lifestyle modifications to prevent and control hypertension. 2. Recommendations on obesity
and weight loss. Canadian Hypertension Society, Canadian Coalition for High Blood Pressure
Prevention and Control, Laboratory Centre for Disease Control at Health Canada, Heart and
280
Hypertension (partial update)
Reference list
Stroke Foundation of Canada. Canadian Medical Association Journal. 1999; 160(9
Suppl):S7-12.
356 Leren P, Helgeland A. Oslo Hypertension Study. Drugs. 1986; 31 Suppl 1:41-45.
357 Levin ML, Goldstein H, Gerhardt PR. Cancer and tobacco smoking; a preliminary
report. JAMA. 1950; 143(4):336-338.
358 Levine DM, Cohen JD, Dustan HP, Falkner B, Flora JA, Lefebvre RC, Morisky DE,
Oberman A, Pickering TG, Roccella EJ, . Behavior changes and the prevention of high blood
pressure. Workshop II. AHA Prevention Conference III. Behavior change and compliance:
keys to improving cardiovascular health. Circulation. 1993; 88(3):1387-1390.
359 Levinson PD, Khatri IM, Freis ED. Persistence of normal BP after withdrawal of drug
treatment in mild hypertension. Archives of Internal Medicine. 1982; 142(13):2265-2268.
360 Levy A, Lightman SL. Diagnosis and management of pituitary tumours. BMJ. 1994;
308(6936):1087-1091.
361 Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific relevance of
usual blood pressure to vascular mortality: a meta-analysis of individual data for one million
adults in 61 prospective studies. Lancet. 2002; 360(9349):1903-1913.
362 Lewis EJ, Hunsicker LG, Clarke WR, Berl T, Pohl MA, Lewis JB, Ritz E, Atkins RC,
Rohde R, Raz I. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in
patients with nephropathy due to type 2 diabetes. New England Journal of Medicine. 2001;
345(12):851-860.
363 Li Y, Boggia J, Thijs L, Hansen TW, Kikuya M, Bjorklund-Bodegard K, Richart T,
Ohkubo T, Kuznetsova T, Torp-Pedersen C, Lind L, Ibsen H, Imai Y, Wang J, Sandoya E,
O'Brien E, Staessen JA, International Database on Ambulatory Blood Pressure Monitoring in
relation to Cardiovascular Outcomes Investigators. Is blood pressure during the night more
predictive of cardiovascular outcome than during the day? Blood Pressure Monitoring. 2008;
13(3):145-147.
364 Lim PO, Rodgers P, Cardale K, Watson AD, MacDonald TM. Potentially high
prevalence of primary aldosteronism in a primary-care population. Lancet. 1999;
353(9146):40.
365 Lind L, Lithell H, Pollare T, Ljunghall S. Blood pressure response during long-term
treatment with magnesium is dependent on magnesium status. A double-blind, placebocontrolled study in essential hypertension and in subjects with high-normal blood pressure.
American Journal of Hypertension. 1991; 4(8):674-679.
366 Linden W, Chambers L. Clinical effectiveness of non-drug treatment for hypertension:
a meta-analysis. Annals of Behavioral Medicine. 1994; 16:35-45.
367 Linden W, Lenz JW, Con AH. Individualized stress management for primary
hypertension: a randomized trial. Archives of Internal Medicine. 2001; 161(8):1071-1080.
368 Lindholm LH, Hansson L, Ekbom T, Dahlöf B, Lanke J, Linjer E, et al. Comparison
of antihypertensive treatments in preventing cardiovascular events in elderly diabetic patients:
results from the Swedish trial in old patients with hypertension -2. Journal of Hypertension.
2000; 18:1671-1675.
369 Lindholm LH, Ibsen H, Borch-Johnsen K, Olsen MH, Wachtell K, Dahlof B,
Devereux RB, Beevers G, de Faire U, Fyhrquist F, Julius S, Kjeldsen SE, Kristianson K,
Lederballe-Pedersen O, Nieminen MS, Omvik P, Oparil S, Wedel H, Aurup P, Edelman JM,
Snapinn S, For the LIFE study group. Risk of new-onset diabetes in the Losartan Intervention
For Endpoint reduction in hypertension study. Journal of Hypertension. 2002; 20(9):18791886.
370 Lindholm LH, Ibsen H, Dahlöf B, Devereux RB, Beevers G, de Faire U, et al.
Cardiovascular morbidity and mortality in patients with diabetes in the Losartan Intervention
For Endpoint reduction in hypertension study (LIFE): a randomised trial against Atenolol.
Lancet. 2002; 359:1004-1010.
281
Hypertension (partial update)
Reference list
371 Lithell H, Hansson L, Skoog I, Elmfeldt D, Hofman A, Olofsson B, Trenkwalder P,
Zanchetti A. The Study on Cognition and Prognosis in the Elderly (SCOPE): principal results
of a randomized double-blind intervention trial. Journal of Hypertension. 2003; 21(5):875886.
372 Little P, Barnett J, Barnsley L, Marjoram J, Fitzgerald-Barron A, Mant D. Comparison
of agreement between different measures of blood pressure in primary care and daytime
ambulatory blood pressure. BMJ. 2002; 325(7358):254.
373 Liu L, Wang JG, Gong L, Liu G, Staessen JA. Comparison of active treatment and
placebo in older Chinese patients with isolated systolic hypertension. Systolic Hypertension in
China (Syst-China) Collaborative Group. Journal of Hypertension. 1998; 16(12 Pt 1):18231829.
374 Lorgelly P, Siatis I, Brooks A, Slinn B, Millar-Craig MW, Donnelly R, Manning G. Is
ambulatory blood pressure monitoring cost-effective in the routine surveillance of treated
hypertension patients in primary care? British Journal of General Practice. 2003; 53:794-796.
375 Luengo-Fernandez R, Gray AM, Rothwell PM. Costs of stroke using patient-level
data: a critical review of the literature. Stroke. 2009; 40(2):e18-e23.
376 Luft FC, Morris CD, Weinburger MH. Compliance to a low sodium diet. American
Journal of Clinical Nutrition. 1997; 65:S698-S703.
377 Luscher TF, Vetter H, Siegenthaler W, Vetter W. Compliance in hypertension: facts
and concepts. Journal of Hypertension - Supplement. 1985; 3(1):S3-S9.
378 Lyle RM. Does baseline serum total calcium level influence the blood pressure
response to calcium supplementation? A double-blind study. Netherlands Journal of
Medicine. 1992; 41(1-2):48-55.
379 Macmahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J,
Dyer A, Stamler J. Blood pressure, stroke, and coronary heart disease. Part 1, Prolonged
differences in blood pressure: prospective observational studies corrected for the regression
dilution bias. Lancet. 1990; 335(8692):765-774.
380 MacMahon SW, Macdonald GJ, Bernstein L, Andrews G, Blacket RB. A randomized
controlled trial of weight reduction and metoprolol in the treatment of hypertension in young
overweight patients. Clinical and Experimental Pharmacology and Physiology. 1985;
12(3):267-271.
381 MacMahon SW, Macdonald GJ, Bernstein L, Andrews G, Blacket RB. Comparison of
weight reduction with metoprolol in treatment of hypertension in young overweight patients.
Lancet. 1985; 1(8440):1233-1236.
382 Maheswaran R, Beevers M, Beevers DG. Effectiveness of advice to reduce alcohol
consumption in hypertensive patients. Hypertension. 1992; 19(1):79-84.
383 Mahmud A, Mahgoub M, Hall M, Feely J. Does aldosterone-to-renin ratio predict the
antihypertensive effect of the aldosterone antagonist spironolactone? American Journal of
Hypertension. 2005; 18(12):1631-1635.
384 Malacco E, Mancia G, Rappelli A, Menotti A, Zuccaro MS, Coppini A, Shell I.
Treatment of isolated systolic hypertension: the SHELL study results. Blood Pressure. 2003;
12(3):160-167.
385 Mallion JM, Genes N, Vaur L, Clerson P, Vaisse B, Bobrie G, Chatellier G. Blood
pressure levels, risk factors and antihypertensive treatments: lessons from the SHEAF study.
Journal of Human Hypertension. 2001; 15(12):841-848.
386 Mancia G, Omboni S, Parati G, Trazzi S, Mutti E. Limited reproducibility of hourly
blood pressure values obtained by ambulatory blood pressure monitoring: Implications for
studies on antihypertensive drugs. Journal of Hypertension. 1992; 10(12):1531-1535.
387 Mancia G, Parati G. Office compared with ambulatory blood pressure in assessing
response to anti hypertensive treatment: a meta-analysis. Journal of Hypertension. 2004;
22(3):435-445.
282
Hypertension (partial update)
Reference list
388 Mandle CL, Jacobs SC, Arcari PM, Domar AD. The efficacy of relaxation response
interventions with adult patients: a review of the literature. Journal of Cardiovascular
Nursing. 1996; 10(3):4-26.
389 Mansoor GA, McCabe EJ, White WB. Long-term reproducibility of ambulatory blood
pressure. Journal of Hypertension. 1994; 12(6):703-708.
390 Mar J, Pastor R, Abasolo R, Ruiz de GR. Ambulatory blood pressure monitoring and
diagnostic errors in hypertension: a Bayesian approach. Medical Decision Making. 1998;
18(4):429-435.
391 Martin JE, Dubbert PM, Cushman WC. Controlled trial of aerobic exercise in
hypertension. Circulation. 1990; 81(5):1560-1567.
392 Mason, J. and Eccles, M. Guideline Recommendation an Evidence Grading (GREG):
a New Grading Method for Clinical Guideline Development Groups. Volume Report 109.
Newcastle upon Tyne: Centre for Health Services Research, University of Newcastle upon
Tyne, 2003.
393 Mason J, Eccles M, Freemantle N, Drummond M. A framework for incorporating
cost-effectiveness in evidence-based clinical practice guidelines. Health Policy. 1999;
47(1):37-52.
394 Materson BJ, Reda DJ, Cushman WC, Massie BM, Freis ED, Kochar MS, Hamburger
RJ, Fye C, Lakshman R, Gottdiener J. Single-drug therapy for hypertension in men. A
comparison of six antihypertensive agents with placebo. The Department of Veterans Affairs
Cooperative Study Group on Antihypertensive Agents. New England Journal of Medicine.
1993; 328(13):914-921.
395 Maxwell MH, Kushiro T, Dornfeld LP, Tuck ML, Waks AU. BP changes in obese
hypertensive subjects during rapid weight loss. Comparison of restricted v unchanged salt
intake. Archives of Internal Medicine. 1984; 144(8):1581-1584.
396 McCarron DA, Morris CD. Blood pressure response to oral calcium in persons with
mild to moderate hypertension. A randomized, double-blind, placebo-controlled, crossover
trial. Annals of Internal Medicine. 1985; 103(6 ( Pt 1)):825-831.
397 McGrady A. Effects of group relaxation training and thermal biofeedback on blood
pressure and related physiological and psychological variables in essential hypertension.
Biofeedback and Self Regulation. 1994; 19(1):51-66.
398 McGrady AV, Yonker R, Tan SY, Fine TH, Woerner M. The effect of biofeedbackassisted relaxation training on blood pressure and selected biochemical parameters in patients
with essential hypertension. Biofeedback and Self Regulation. 1981; 6(3):343-353.
399 McKeigue PM, Marmot MG, Syndercombe Court YD, Cottier DE, Rahman S,
Riemersma RA. Diabetes, hyperinsulinaemia, and coronary risk factors in Bangladeshis in
east London. British Heart Journal. 1988; 60(5):390-396.
400 McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin
resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet.
1991; 337(8738):382-386.
401 McManus RJ, Mant J, Bray EP, Holder R, Jones MI, Greenfield S, Kaambwa B,
Banting M, Bryan S, Little P, Williams B, Hobbs FD. Telemonitoring and self-management
in the control of hypertension (TASMINH2): a randomised controlled trial. Lancet. 2010;
376(9736):163-172.
402 Medical Research Council Working Party. MRC trial of treatment of mild
hypertension: principal results. BMJ. 1985; 291(6488):97-104.
403 Melsop KA, Boothroyd DB, Hlatky MA. Quality of life and time trade-off utility
measures in patients with coronary artery disease. American Heart Journal. 2003; 145(1):3641.
283
Hypertension (partial update)
Reference list
404 Mesquita-Bastos J, Bertoquini S, Polonia J. Cardiovascular prognostic value of
ambulatory blood pressure monitoring in a Portuguese hypertensive population followed up
for 8.2 years. Blood Pressure Monitoring. 2010; 15(5):240-246.
405 Metoki H, Ohkubo T, Kikuya M, Asayama K, Obara T, Hara A, Hirose T, Hashimoto
J, Totsune K, Hoshi H, Satoh H, Imai Y. Prognostic significance of night-time, early morning,
and daytime blood pressures on the risk of cerebrovascular and cardiovascular mortality: the
Ohasama Study. Journal of Hypertension. 2006; 24(9):1841-1848.
406 Metz JA, Stern JS, Kris-Etherton P, Reusser ME, Morris CD, Hatton DC, Oparil S,
Haynes RB, Resnick LM, Pi-Sunyer FX, Clark S, Chester L, McMahon M, Snyder GW,
McCarron DA. A randomized trial of improved weight loss with a prepared meal plan in
overweight and obese patients: impact on cardiovascular risk reduction. Archives of Internal
Medicine. 2000; 160(14):2150-2158.
407 Midgley JP, Matthew AG, Greenwood CM, Logan AG. Effect of reduced dietary
sodium on blood pressure: a meta-analysis of randomized controlled trials. JAMA. 1996;
275(20):1590-1597.
408 Miller ER, III, Erlinger TP, Young DR, Jehn M, Charleston J, Rhodes D, Wasan SK,
Appel LJ. Results of the Diet, Exercise, and Weight Loss Intervention Trial (DEW-IT).
Hypertension. 2002; 40(5):612-618.
409 Miller JM, Miller JM. Diabetes mellitus and hypertension in black and white
populations. Southern Medical Journal. 1986; 79(10):1229.
410 Mills CA, Porter MM. Tobacco smoking habits and cancer of the mouth and
respiratory system. Cancer Research. 1950; 10(9):539-542.
411 Mitchell A, Haynes RB, Adsett CA, Bellissimo A, Wilczynski N. The likelihood of
remaining normotensive following antihypertensive drug withdrawal. Journal of General
Internal Medicine. 1989; 4(3):221-225.
412 Montgomery AA, Fahey T, Ben-Shlomo Y, Harding J. The influence of absolute
cardiovascular risk, patient utilities, and costs on the decision to treat hypertension: a Markov
decision analysis. Journal of Hypertension. 2003; 21(9):1753-1759.
413 Morgan T, Hopper J, Anderson A, Carricks L, Jones E, Johns J. Can drug therapy be
stopped in elderly hypertensive patients? Cardiology in the Elderly. 1994; 2:119-125.
414 Morris CD, Karanja N, McCarron DA. Dietary versus supplemental calcium to reduce
blood pressure. Clinical Research. 1988; 36:A139.
415 Mulrow CD, Chiquette E, Angel L, Cornell J, Summerbell C, Anagnostelis B, Grimm
R, Jr., Brand MB. Dieting to reduce body weight for controlling hypertension in adults.
Cochrane Database of Systematic Reviews. 2000;(2):CD000484.
416 Murakami S, Otsuka K, Kubo Y, Shinagawa M, Yamanaka T, Ohkawa S-I, Kitaura Y.
Repeated ambulatory monitoring reveals a Monday morning surge in blood pressure in a
community-dwelling population. American Journal of Hypertension. 2004; 17(12):11791183.
417 Murchie P, Campbell NC, Ritchie LD, Simpson JA, Thain J. Secondary prevention
clinics for coronary heart disease: four year follow up of a randomised controlled trial in
primary care. BMJ. 2003; 326(7380):84.
418 Murugesan R, Govindarajulu N, Bera TK. Effect of selected yogic practices on the
management of hypertension. Indian Journal of Physiology and Pharmacology. 2000;
44(2):207-210.
419 Musini VM, Tejani AM, Bassett K, Wright JM. Pharmacotherapy for hypertension in
the elderly. Cochrane Database of Systematic Reviews. 2009; Issue 4:CD000028.
420 Musso NR, Vergassola C, Barone C, Lotti G. Ambulatory blood pressure monitoring:
how reproducible is it? American Journal of Hypertension. 1997; 10(8):936-939.
421 Myers MG, Reeves RA, Oh PI, Joyner CD. Overtreatment of hypertension in the
community? American Journal of Hypertension. 1996; 9(5):419-425.
284
Hypertension (partial update)
Reference list
422 National Centre for Social Research and University College London.Department of
Epidemiology and Public Health.Health Survey for England, 2006 [Computer File]. 3rd ed.
Colchester: Essex: UK Data Archive [distributor]; 2010
423 National Clinical Guideline Centre. Unstable Angina and NSTEMI: the Early
Management of Unstable Angina and Non-ST-Segment-Elevation Myocardial Infarction.
(CG94). London: Royal College of Physicians, 2010.
424 National Clinical Guideline Centre for Acute and Chronic Conditions. Management of
Stable Angina: NICE Guideline: Draft for Consultation, December 2010. 2010. London,
National Clinical Guideline Centre for Acute and Chronic Conditions.
http://www.nice.org.uk/nicemedia/live/11878/52141/52141.pdf
425 National Collaborating Centre for Chronic Conditions.Hypertension: Management in
Adults in Primary Care: Pharmacological Update. London: Royal College of Physicians;
2006. Available from: http://guidance.nice.org.uk/CG34
426 National Collaborating Centre for Primary Care.Medicines Adherence - Involving
Patients in Decisions About Prescribed Medicines and Supporting Adherence. National
Clinical Guideline Number 76. London: National Collaborating Centre for Primary Care and
Royal College of General Practitioners; 2009. Available from:
http://guidance.nice.org.uk/CG76
427 National Institute for Health and Clinical Excellence. Guide to the Methods of
Technology Appraisal. London: National Institute for Health and Clinical Excellence, 2008.
428 National Institute for Health and Clinical Excellence. Lipid Modification:
Cardiovascular Risk Assessment and the Modification of Blood Lipids for the Primary and
Secondary Prevention of Cardiovascular Disease. London: National Institute for Health and
Clinical Excellence, 2008.
429 National Institute for Health and Clinical Excellence. Social Value Judgements.
Principles for the development of NICE guidance. Second edition. Available from:
http://www.nice.org.uk/media/C18/30/SVJ2PUBLICATION2008.pdf Last accessed on: 15
November 2010.
430 National Institute for Health and Clinical Excellence.The Guidelines Manual. London:
National Institute for Health and Clinical Excellence; 2009. Available from:
http://www.nice.org.uk/media/5F2/44/The_guidelines_manual_2009_-_All_chapters.pdf
431 National Institute for Health and Clinical Excellence. Chest Pain of Recent Onset:
Assessment and Diagnosis of Recent Onset Chest Pain or Discomfort of Suspected Cardiac
Origin. London: National Institute for Health and Clinical Excellence, 2010.
432 National Institute for Health and Clinical Excellence. Clopidogrel and ModifiedRelease Dipyridamole for the Prevention of Occlusive Vascular Events: Review of NICE
Technology Appraisal Guidance 90. London: National Institute for Health and Clinical
Excellence, 2010.
433 Nelson M. Author's reply to two responses. BMJ. 2002;
434 Nelson M, Reid C, Krum H, McNeil J. A systematic review of predictors of
maintenance of normotension after withdrawal of antihypertensive drugs. American Journal
of Hypertension. 2001; 14(2):98-105.
435 Nelson MR, Reid CM, Krum H, Muir T, Ryan P, McNeil JJ. Predictors of
normotension on withdrawal of antihypertensive drugs in elderly patients: prospective study
in second Australian national blood pressure study cohort. BMJ. 2002; 325(7368):815.
436 Newcastle Guideline Development and Research Unit. Hypertension: Management of
Hypertension in Adults in Primary Care. London: National Institute for Clinical Excellence,
2004.
437 Newell-Price J, Trainer P, Besser M, Grossman A. The diagnosis and differential
diagnosis of Cushing's syndrome and pseudo-Cushing's states. Endocrinology Review. 1998;
19(5):647-672.
285
Hypertension (partial update)
Reference list
438 Niiranen TJ, Hanninen MR, Johansson J, Reunanen A, Jula AM. Home-measured
blood pressure is a stronger predictor of cardiovascular risk than office blood pressure: the
Finn-Home study. Hypertension. 2010; 55(6):1346-1351.
439 Niiranen TJ, Kantola IM, Vesalainen R, Johansson J, Ruuska MJ. A comparison of
home measurement and ambulatory monitoring of blood pressure in the adjustment of
antihypertensive treatment. American Journal of Hypertension. 2006; 19(5):468-474.
440 North of England Evidence-based Guidelines Development Project. Primary Care
Management of Secondary Prophylaxis for Patients Who Have Experienced a Myocardial
Infarction: Drug Treatment, Cardiac Rehabilitation and Dietary Manipulation. Newcastle:
University of Newcastle, Centre for Health Services Research, 2001.
441 North of England Hypertension Guideline Development Group. Essential
Hypertension: Managing Adult Patients in Primary Care. Newcastle upon Tyne: Cetnre for
Health Services Research, 2004.
442 Nowson C, Morgan T. Effect of calcium carbonate on blood pressure in normotensive
and hypertensive people. Hypertension. 1989; 13(6 Pt 1):630-639.
443 Nowson CA, Morgan TO. Magnesium supplementation in mild hypertensive patients
on a moderately low sodium diet. Clinical and Experimental Pharmacology and Physiology.
1989; 16(4):299-302.
444 Nurminen ML, Niittynen L, Korpela R, Vapaatalo H. Coffee, caffeine and blood
pressure: a critical review. European Journal of Clinical Nutrition. 1999; 53(11):831-839.
445 O'Brien E. Replacing the mercury sphygmomanometer. Requires clinicians to demand
better automated devices. BMJ. 2000; 320(7238):815-816.
446 O'Brien E. State of the market for devices for blood pressure measurement. Blood
Pressure Monitoring. 2001; 6(6):281-286.
447 O'Brien E, Atkins N. A comparison of the British Hypertension Society and
Association for the Advancement of Medical Instrumentation protocols for validating blood
pressure measuring devices: can the two be reconciled? Journal of Hypertension. 1994;
12(9):1089-1094.
448 O'Brien E, Beevers G, Lip G. ABC of hypertension: Blood pressure measurement.
Part III. BMJ. 2001; 322(7294):1110-1114.
449 O'Brien E, Beevers G, Lip GY. ABC of hypertension: Blood pressure measurement.
Part IV-automated sphygmomanometry: self blood pressure measurement. BMJ. 2001;
322(7295):1167-1170.
450 O'Brien E, Coats A, Owens P, Petrie J, Padfield PL, Littler WA, de SM, Mee F. Use
and interpretation of ambulatory blood pressure monitoring: recommendations of the British
hypertension society. BMJ. 2000; 320(7242):1128-1134.
451 O'Brien E, Petrie J, Littler W, de SM, Padfield PL, Altman DG, Bland M, Coats A,
Atkins N. An outline of the revised British Hypertension Society protocol for the evaluation
of blood pressure measuring devices. Journal of Hypertension. 1993; 11(6):677-679.
452 O'Brien E, Waeber B, Parati G, Staessen J, Myers MG. Blood pressure measuring
devices: recommendations of the European Society of Hypertension. BMJ. 2001;
322(7285):531-536.
453 O'Malley K, McCormack P, O'Brien ET. Isolated systolic hypertension: data from the
European Working Party on High Blood Pressure in the Elderly. Journal of Hypertension Supplement. 1988; 6(1):S105-S108.
454 Obara F, Saitoh S, Takagi S, Shimamoto K. Influence of hypertension on the
incidence of cardiovascular disease in two rural communities in Japan: the Tanno-Sobetsu
[corrected] study. Hypertension Research - Clinical and Experimental. 2007; 30(8):677-682.
455 Obel AO. Placebo-controlled trial of potassium supplements in black patients with
mild essential hypertension. Journal of Cardiovascular Pharmacology. 1989; 14(2):294-296.
286
Hypertension (partial update)
Reference list
456 Octavio JA, Contreras J, Amair P, Octavio B, Fabiano D, Moleiro F, Omboni S,
Groppelli A, Bilo G, Mancia G, Parati G. Time-weighted vs. conventional quantification of
24-h average systolic and diastolic ambulatory blood pressures. Journal of Hypertension.
2010; 28(3):459-464.
457 Office for National Statistics. Mortality Statistcs: Cause: Review of the Registrar
General on Deaths by Cause, Sex and Age in England and Wales, 2000. London: Office of
National Statistics, 2001.
458 Office for National Statistics. Life Tables (Online Edition). 2011. London, Office for
National Statistics.
459 Office for National Statistics (ONS). England and Wales, Interim Life Tables, 1980-82
to 2007-09. Available from: http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=14459
Last accessed on: 21 October 2010.
460 Office for National Statistics (ONS). Mortality statistics: deaths registered in 2008.
Available from: http://www.statistics.gov.uk/statbase/product.asp?vlnk=15096 Last accessed
on: 21 October 2010.
461 Ogedegbe G, Pickering TG, Clemow L, Chaplin W, Spruill TM, Albanese GM,
Eguchi K, Burg M, Gerin W. The misdiagnosis of hypertension: The role of patient anxiety.
Archives of Internal Medicine. 2008; 168(22):2459-2465.
462 Ogihara T, Saruta T, Rakugi H, Fujimoto A, Ueshima K, Yasuno S, Oba K, Takeda K,
Higaki J, Nakao K. Relationship between the achieved blood pressure and the incidence of
cardiovascular events in Japanese hypertensive patients with complications: A sub-analysis of
the CASE-J trial. Hypertension Research. 2009; 32(4):248-254.
463 Ogihara T, Saruta T, Rakugi H, Matsuoka H, Shimamoto K, Shimada K, Imai Y,
Kikuchi K, Ito S, Eto T, Kimura G, Imaizumi T, Takishita S, Ueshima H, Valsartan in Elderly
Isolated Systolic Hypertension Study Group. Target blood pressure for treatment of isolated
systolic hypertension in the elderly: valsartan in elderly isolated systolic hypertension study.
Hypertension. 2010; 56(2):196-202.
464 Ohkubo T, Asayama K, Kikuya M, Metoki H, Hoshi H, Hashimoto J, Totsune K,
Satoh H, Imai Y, Ohasama S. How many times should blood pressure be measured at home
for better prediction of stroke risk? Ten-year follow-up results from the Ohasama study.
Journal of Hypertension. 2004; 22(6):1099-1104.
465 Ohkubo T, Imai Y, Tsuji I, Nagai K, Watanabe N, Minami N, Itoh O, Bando T,
Sakuma M, Fukao A, Satoh H, Hisamichi S, Abe K. Prediction of mortality by ambulatory
blood pressure monitoring versus screening blood pressure measurements: a pilot study in
Ohasama. Journal of Hypertension. 1997; 15(4):357-364.
466 Okayama A, Kadowaki T, Okamura T, Hayakawa T, Ueshima H, NIPPON DATA.
Age-specific effects of systolic and diastolic blood pressures on mortality due to
cardiovascular diseases among Japanese men (NIPPON DATA80). Journal of Hypertension.
2006; 24(3):459-462.
467 Onusko E. Diagnosing secondary hypertension. American Family Physician. 2003;
67(1):67-74.
468 Organisation for Economic Co-operation and Development (OECD). OECD Stat
Extracts: purchasing power parities for GDP. Available from:
http://stats.oecd.org/Index.aspx?datasetcode=SNA_TABLE4 Last accessed on: 10 November
2010.
469 Orth DN. Cushing's syndrome. New England Journal of Medicine. 1995; 332(12):791803.
470 Overlack A, Conrad H, Stumpe KO. The influence of oral potassium
citrate/bicarbonate on blood pressure in essential hypertension during unrestricted salt intake.
Klinische Wochenschrift. 1991; 69 Suppl 25:79-83.
287
Hypertension (partial update)
Reference list
471 Pagliari CP, Grimshaw JM, Eccles M. The potential influence of small group
processes on guideline development. Journal of Evaluation in Clinical Practice. 2001;
7(2):165-173.
472 Palatini P, Julius S, Collatina S, Rappelli S, Staessen J, Pessina AC. Optimizing the
assessment of the elderly patient with borderline hypertension: the Hypertension and
Ambulatory Recording in the OLD (HAROLD) study. Aging. 1997; 9(5):365-371.
473 Palatini P, Mormino P, Canali C, Santonastaso M, De VG, Zanata G, Pessina AC.
Factors affecting ambulatory blood pressure reproducibility. Results of the HARVEST Trial.
Hypertension and Ambulatory Recording Venetia Study. Hypertension. 1994; 23(2):211-216.
474 Palmer, S, Sculpher, M, Philips, Z, Robinson, M, and et al. A Cost Effectiveness
Model Comparing Alternative Management Strategies for the Use of Glycoprotein IIB/IIIA
Anatagonists in Non-ST-Elevation Acute Conorary Syndrome. York: Center for Health
Economics, 2002.
475 Paolisso G, Di MG, Cozzolino D, Salvatore T, D'Amore A, Lama D, Varricchio M,
D'Onofrio F. Chronic magnesium administration enhances oxidative glucose metabolism in
thiazide treated hypertensive patients. American Journal of Hypertension. 1992; 5(10):681686.
476 Parati G, Omboni S, Albini F, Piantoni L, Giuliano A, Revera M, Illyes M, Mancia G,
TeleBPCare Study Group. Home blood pressure telemonitoring improves hypertension
control in general practice. The TeleBPCare study. Journal of Hypertension. 2009; 27(1):198203.
477 Patel C, Marmot M. Can general practitioners use training in relaxation and
management of stress to reduce mild hypertension? BMJ. 1988; 296(6614):21-24.
478 Patel C, Marmot MG, Terry DJ. Controlled trial of biofeedback-aided behavioural
methods in reducing mild hypertension. BMJ. 1981; 282(6281):2005-2008.
479 Patel C, Marmot MG, Terry DJ, Carruthers M, Hunt B, Patel M. Trial of relaxation in
reducing coronary risk: four year follow up. BMJ. 1985; 290(6475):1103-1106.
480 Payne R. Excel worksheet for calculating cardiovascular risk. Available from:
http://cvrisk.mvm.ed.ac.uk/calculator/excelcalc.htm
481 Pepine CJ, Handberg EM, Cooper-DeHoff RM, Marks RG, Kowey P, Messerli FH,
Mancia G, Cangiano JL, Garcia-Barreto D, Keltai M, Erdine S, Bristol HA, Kolb HR, Bakris
GL, Cohen JD, Parmley WW. A calcium antagonist vs a non-calcium antagonist hypertension
treatment strategy for patients with coronary artery disease. The International VerapamilTrandolapril Study (INVEST): a randomized controlled trial. JAMA. 2003; 290(21):28052816.
482 Perloff D, Grim C, Flack J, Frohlich ED, Hill M, McDonald M, Morgenstern BZ.
Human blood pressure determination by sphygmomanometry. Circulation. 1993; 88(5 Pt
1):2460-2470.
483 Perry HM, Davis BR, Price TR, Applegate WB, Fields WS, Guralnik JM, Kuller L,
Pressel S, Stamler J, Probsfieldt JL. Effect of treating isolated systolic hypertension on the
risk of developing various types and subtypes of stroke: the Systolic Hypertension in the
Elderly Program (SHEP). JAMA. 2000; 284:465-471.
484 Perry HM, McDonald RH, Hulley SB, Smith WM, Furberg CD, Greenlick MR.
Systolic Hypertension in the Elderly Program. Pilot Study (SHEP-PS): morbidity and
mortality experience. Journal of Hypertension. 1986; 4(6):S21-S23.
485 Perry HM, Jr., Smith WM, McDonald RH, Black D, Cutler JA, Furberg CD,
Greenlick MR, Kuller LH, Schnaper HW, Schoenberger JA, . Morbidity and mortality in the
Systolic Hypertension in the Elderly Program (SHEP) pilot study. Stroke. 1989; 20(1):4-13.
486 Personal Social Services Research Unit. Unit costs of health and social care 2010.
Available from: http://www.pssru.ac.uk/uc/uc2010contents.htm Last accessed on: 23
December 10 A.D.
288
Hypertension (partial update)
Reference list
487 Petersen LJ, Rudnicki M, Hojsted J. Long-term oral calcium supplementation reduces
diastolic blood pressure in end stage renal disease. A randomized, double-blind, placebo
controlled study. International Journal of Artificial Organs. 1994; 17(1):37-40.
488 Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. Smoking, smoking cessation,
and lung cancer in the UK since 1950: combination of national statistics with two case-control
studies. BMJ. 2000; 321(7257):323-329.
489 Petrella RJ. How effective is exercise training for the treatment of hypertension?
Clinical Journal of Sport Medicine. 1998; 8(3):224-231.
490 Pickering T. Recommendations for the use of home (self) and ambulatory blood
pressure monitoring. American Society of Hypertension Ad Hoc Panel. American Journal of
Hypertension. 1996; 9(1):1-11.
491 Pickering T, Schwartz J, Verdecchia P, Imai Y, Kario K, Eguchi K, Pierdomenico S,
Ohkubo T, Wing L. Prediction of strokes versus cardiac events by ambulatory monitoring of
blood pressure: results from an international database. Blood Pressure Monitoring. 2007;
12(6):397-399.
492 Piller LB, Ford CE, Davis BR, Nwachuku C, Black HR, Oparil S, Retta TM,
Probstfield JL, ALLHAT Collaborative Research Group. Incidence and predictors of
angioedema in elderly hypertensive patients at high risk for cardiovascular disease: a report
from the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial
(ALLHAT). Journal of Clinical Hypertension. 2006; 8(9):649-656.
493 Plante GE, Dessurault DL. Hypertension in elderly patients. A comparative study
between indapamide and hydrochlorothiazide. American Journal of Medicine. 1988;
84(1B):98-103.
494 Plante GE, Robillard C. Indapamide in the treatment of essential arterial hypertension:
results of a controlled study. Current Medical Research and Opinion. 1983; 8 Suppl 3:59-66.
495 Poppitt SD, Keogh GF, Prentice AM, Williams DE, Sonnemans HM, Valk EE,
Robinson E, Wareham NJ. Long-term effects of ad libitum low-fat, high-carbohydrate diets
on body weight and serum lipids in overweight subjects with metabolic syndrome. American
Journal of Clinical Nutrition. 2002; 75(1):11-20.
496 Poulter NR, Khaw KT, Hopwood BE, Mugambi M, Peart WS, Rose G, Sever PS. The
Kenyan Luo migration study: observations on the initiation of a rise in blood pressure. BMJ.
1990; 300(6730):967-972.
497 Power C, Rodgers B, Hope S. U-shaped relation for alcohol consumption and health in
early adulthood and implications for mortality. Lancet. 1998; 352(9131):877.
498 Prisco D, Paniccia R, Bandinelli B, Filippini M, Francalanci I, Giusti B, Giurlani L,
Gensini GF, Abbate R, Neri Serneri GG. Effect of medium-term supplementation with a
moderate dose of n-3 polyunsaturated fatty acids on blood pressure in mild hypertensive
patients. Thrombosis Research. 1998; 91(3):105-112.
499 Pritchard DA, Hyndman J, Taba F. Nutritional counselling in general practice: a cost
effective analysis. Journal of Epidemiology and Community Health. 1999; 53(5):311-316.
500 PROGRESS Collaborative Group. Randomised trial of a perindopril-based bloodpressure-lowering regimen among 6,105 individuals with previous stroke or transient
ischaemic attack. Lancet. 2001; 358(9287):1033-1041.
501 Psaty BM, Smith NL, Siscovick DS, Koepsell TD, Weiss NS, Heckbert SR, Lemaitre
RN, Wagner EH, Furberg CD. Health outcomes associated with antihypertensive therapies
used as first-line agents. A systematic review and meta-analysis. JAMA. 1997; 277(9):739745.
502 Puddey IB, Parker M, Beilin LJ, Vandongen R, Masarei JR. Effects of alcohol and
caloric restrictions on blood pressure and serum lipids in overweight men. Hypertension.
1992; 20(4):533-541.
289
Hypertension (partial update)
Reference list
503 Rakic D, Rumboldt Z, Bagatin J, Polic S. Effects of four antihypertensive
monotherapies on cardiac mass and function in hypertensive patients with left ventricular
hypertrophy: randomized prospective study. Croatian Medical Journal. 2002; 43(6):672-679.
504 Raleigh VS. Diabetes and hypertension in britain's ethnic minorities: implications for
the future of renal services. BMJ. 1997; 314(7075):209.
505 Ramsay L, Williams B, Johnston G, MacGregor G, Poston L, Potter J, Poulter N,
Russell G. Guidelines for management of hypertension: report of the third working party of
the British Hypertension Society. Journal of Human Hypertension. 1999; 13(9):569-592.
506 Redon J, Campos C, Narciso ML, Rodicio JL, Pascual JM, Ruilope LM. Prognostic
value of ambulatory blood pressure monitoring in refractory hypertension: a prospective
study. Hypertension. 1998; 31(2):712-718.
507 Reims HM, Oparil S, Kjeldsen SE, Devereux RB, Julius S, Brady WE, Fyhrquist F,
Ibsen H, Lindholm LH, Omvik P, Wedel H, Beevers G, de Faire U, Kristianson K,
Lederballe-Pedersen O, Nieminen MS, Dahlof B, LIFE Study Group. Losartan benefits over
atenolol in non-smoking hypertensive patients with left ventricular hypertrophy: the LIFE
study. Blood Pressure. 2004; 13(6):376-384.
508 Reisin E, Abel R, Modan M, Silverberg DS, Eliahou HE, Modan B. Effect of weight
loss without salt restriction on the reduction of blood pressure in overweight hypertensive
patients. New England Journal of Medicine. 1978; 298(1):1-6.
509 Richards AM, Nicholls MG, Espiner EA, Ikram H, Turner JG, Brownlie BE.
Hypertension in hypothyroidism: arterial pressure and hormone relationships. Clinical and
Experimental Hypertension Part A: Theory and Practice. 1985; 7(11):1499-1514.
510 Rivas M, Garay RP, Escanero JF, Cia P, Jr., Cia P, Alda JO. Soy milk lowers blood
pressure in men and women with mild to moderate essential hypertension. Journal of
Nutrition. 2002; 132(7):1900-1902.
511 Rodilla E, Costa JA, Perez-Lahiguera F, Baldo E, Gonzalez C, Pascual JM.
Spironolactone and doxazosin treatment in patients with resistant hypertension. Revista
Espanola De Cardiologia. 2009; 62(2):158-166.
512 Rodriguez Roca GC, onso Moreno FJ, Garcia JA, Vega A, Llisterri Caro JL, Barrios
A, V, Segura FA, Clemente = Lirola E et al. Cost-effectiveness of ambulatory blood pressure
monitoring in the follow-up of hypertension. Blood Pressure. 2006; 15(1):27-36.
513 Rogers MW, Probst MM, Gruber JJ, Berger R, Boone JB, Jr. Differential effects of
exercise training intensity on blood pressure and cardiovascular responses to stress in
borderline hypertensive humans. Journal of Hypertension. 1996; 14(11):1369-1375.
514 Rosei EA, Dal Palú C, Leonetti G, Magnani B, Pessina A, Zanchetti A. Clinical results
of the Verapamil in Hypertension and Atherosclerosis Study. Journal of Hypertension. 1997;
15:1337-1344.
515 Rosengren A, Wilhelmsen L, Hagman M, Wedel H. Natural history of myocardial
infarction and angina pectoris in a general population sample of middle-aged men: a 16-year
follow-up of the Primary Prevention Study, Goteborg, Sweden. Journal of Internal Medicine.
1998; 244(6):495-505.
516 Rosmarin PC. Coffee and coronary heart disease: a review. Progress in
Cardiovascular Diseases. 1989; 32(3):239-245.
517 Roth RP, Cantekin EI, Bluestone CD, Welch RM, Cho YW. Nasal decongestant
activity of pseudoephedrine. Annals of Otology, Rhinology and Laryngology. 1977; 86(2 pt.
1):235-242.
518 Royal Pharmaceutical Society of Great Britian Working Party. Partnership in
Medicine Taking: a Consultative Document. London: Royal Pharmaceutical Society of Great
Britain and Mercke Sharpe and Dohme, 1996.
290
Hypertension (partial update)
Reference list
519 Sacks FM, Brown LE, Appel L, Borhani NO, Evans D, Whelton P. Combinations of
potassium, calcium, and magnesium supplements in hypertension. Hypertension. 1995; 26(6
Pt 1):950-956.
520 Sacks FM, Obarzanek E, Windhauser MM, Svetkey LP, Vollmer WM, McCullough
M, Karanja N, Lin PH, Steele P, Proschan MA, . Rationale and design of the Dietary
Approaches to Stop Hypertension trial (DASH). A multicenter controlled-feeding study of
dietary patterns to lower blood pressure. Annals of Epidemiology. 1995; 5(2):108-118.
521 Sairenchi T, Iso H, Irie F, Fukasawa N, Yamagishi K, Kanashiki M, Saito Y, Ota H,
Nose T. Age-specific relationship between blood pressure and the risk of total and
cardiovascular mortality in Japanese men and women. Hypertension Research - Clinical and
Experimental. 2005; 28(11):901-909.
522 Saito I, Kobayashi M, Matsushita Y, Saruta T. Pharmacoeconomical evaluation of
combination therapy for lifetime hypertension treatment in Japan. Journal of the Medical
Association of Japan. 2005; 48(12):574-585.
523 Sakuma M, Imai Y, Tsuji I, Nagai K, Ohkubo T, Watanabe N, Sakuma H, Satoh H,
Hisamichi S. Predictive value of home blood pressure measurement in relation to stroke
morbidity: a population-based pilot study in Ohasama, Japan. Hypertension Research. 1997;
20(3):167-174.
524 Sareli P, Radevski IV, Valtchanova ZP, Libhaber E, Candy GP, Den HE, Libhaber C,
Skudicky D, Wang JG, Staessen JA. Efficacy of different drug classes used to initiate
antihypertensive treatment in black subjects: results of a randomized trial in Johannesburg,
South Africa. Archives of Internal Medicine. 2001; 161(7):965-971.
525 Schein MH, Gavish B, Herz M, Rosner-Kahana D, Naveh P, Knishkowy B, Zlotnikov
E, Ben-Zvi N, Melmed RN. Treating hypertension with a device that slows and regularises
breathing: a randomised, double-blind controlled study. Journal of Human Hypertension.
2001; 15(4):271-278.
526 Schillaci G, Verdecchia P, Porcellati C, Cuccurullo O, Cosco C, Perticone F.
Continuous relation between left ventricular mass and cardiovascular risk in essential
hypertension. Hypertension. 2000; 35(2):580-586.
527 Schillaci G, Verdecchia P, Zampi I, Battistelli M, Bartoccini C, Porcellati C. Noninvasive ambulatory BP monitoring during the night: Randomised comparison of different
reading intervals. Journal of Human Hypertension. 1994; 8(1):23-27.
528 Schrader J, Luders S, Kulschewski A, Hammersen F, Plate K, Berger J, Zidek W,
Dominiak P, Diener HC. Morbidity and Mortality After Stroke, Eprosartan Compared with
Nitrendipine for Secondary Prevention: principal results of a prospective randomized
controlled study (MOSES). Stroke. 2005; 36(6):1218-1226.
529 Schwander B, Gradl B, Zollner Y, Lindgren P, Diener HC, Luders S, Schrader J,
Villar FA, Greiner W, Jonsson B. Cost-utility analysis of eprosartan compared to enalapril in
primary prevention and nitrendipine in secondary prevention in Europe--the HEALTH model.
Value in Health. 2009; 12(6):857-871.
530 Schwartz GL, Turner ST, Moore JH, Sing CF. Effect of time of day on intraindividual
variability in ambulatory blood pressure. American Journal of Hypertension. 2000;
13(11):1203-1209.
531 Schwartz GL, Turner ST, Moore JH, Sing CF. Predictors of interindividual variation
in ambulatory blood pressure and their time or activity dependence. American Journal of
Hypertension. 2000; 13(1 I):52-60.
532 Scientific AdvisoryCommittee on Nutrition. Salt and Health. Norwich: The
Stationery Office, 2003.
533 Seer P, Raeburn JM. Meditation training and essential hypertension: a methodological
study. Journal of Behavioral Medicine. 1980; 3(1):59-71.
291
Hypertension (partial update)
Reference list
534 Sega R, Facchetti R, Bombelli M, Cesana G, Corrao G, Grassi G, Mancia G.
Prognostic value of ambulatory and home blood pressures compared with office blood
pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate
e Loro Associazioni (PAMELA) study. Circulation. 2005; 111(14):1777-1783.
535 Semple PF. Investigation. ABC of Hypertension: Articles From the British Medical
Journal, London: British Medical Asssociation, 1981: 38-41.
536 SHEP Cooperative Research Group. Rationale and design of a randomized clinical
trial on prevention of stroke in isolated systolic hypertension. Journal of Clinical
Epidemiology. 1988; 41(12):1197-1208.
537 SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug
treatment in older persons with isolated systolic hypertension: final results of the Systolic
Hypertension in the Elderly Program (SHEP). JAMA. 1991; 265(24):3255-3264.
538 Sheps SG, Bailey KR, Zachariah PK. Short-term (six hour), ambulatory blood
pressure monitoring. Journal of Human Hypertension. 1994; 8(12):873-878.
539 Shimamoto K, Fujita T, Ito S, Naritomi H, Ogihara T, Shimada K, Tanaka H,
Yoshiike N, -HEALTH Study Committee. Impact of blood pressure control on cardiovascular
events in 26,512 Japanese hypertensive patients: the Japan Hypertension Evaluation with
Angiotensin II Antagonist Losartan Therapy (J-HEALTH) study, a prospective nationwide
observational study. Hypertension Research - Clinical and Experimental. 2008; 31(3):469478.
540 Shimbo D, Kuruvilla S, Haas D, Pickering TG, Schwartz JE, Gerin W. Preventing
misdiagnosis of ambulatory hypertension: algorithm using office and home blood pressures.
Journal of Hypertension. 2009; 27(9):1775-1783.
541 Shinagawa M, Otsuka K, Murakami S, Kubo Y, Cornelissen G, Matsubayashi K,
Yano S, Mitsutake G, Yasaka K, Halberg F. Seven-day (24-h) ambulatory blood pressure
monitoring, self-reported depression and quality of life scores. Blood Pressure Monitoring.
2002; 7(1):69-76.
542 Shulman NB, Hall WD. Renal vascular disease in African-Americans and other racial
minorities. Circulation. 1991; 83(4):1477-1479.
543 Siani A, Strazzullo P, Russo L, Guglielmi S, Iacoviello L, Ferrara LA, Mancini M.
Controlled trial of long term oral potassium supplements in patients with mild hypertension.
BMJ. 1987; 294(6585):1453-1456.
544 Silman AJ, Locke C, Mitchell P, Humpherson P. Evaluation of the effectiveness of a
low sodium diet in the treatment of mild to moderate hypertension. Lancet. 1983;
1(8335):1179-1182.
545 Singh RB, Niaz MA, Bishnoi I, Singh U, Begum R, Rastogi SS. Effect of low energy
diet and weight loss on major risk factors, central obesity and associated disturbances in
patients with essential hypertension. Journal of Human Hypertension. 1995; 9(5):355-362.
546 Sleight P, Redon J, Verdecchia P, Mancia G, Gao P, Fagard R, Schumacher H, Weber
M, Böhm M, Williams B, Pogue J, Koon T, Yusuf S, ONTARGET i. Prognostic value of
blood pressure in patients with high vascular risk in the Ongoing Telmisartan Alone and in
combination with Ramipril Global Endpoint Trial study. Journal of Hypertension. 2009;
27(7):1360-1369.
547 Smith MB, Feldman W. Over-the-counter cold medications. A critical review of
clinical trials between 1950 and 1991. JAMA. 1993; 269(17):2258-2263.
548 Smith WM. Treatment of mild hypertension: results of a ten-year intervention trial.
Circulation Research. 1977; 40(5 Suppl 1):I98-105.
549 Solomon SD, Verma A, Desai A, Hassanein A, Izzo J, Oparil S, Lacourciere Y, Lee J,
Seifu Y, Hilkert RJ, Rocha R, Pitt B, Exforge Intensive Control of Hypertension to Evaluate
Efficacy in Diastolic Dysfunction Investigators. Effect of intensive versus standard blood
292
Hypertension (partial update)
Reference list
pressure lowering on diastolic function in patients with uncontrolled hypertension and
diastolic dysfunction. Hypertension. 2010; 55(2):241-248.
550 Spence JD, Barnett PA, Linden W, Ramsden V, Taenzer P. Lifestyle modifications to
prevent and control hypertension. 7. Recommendations on stress management. Canadian
Hypertension Society, Canadian Coalition for High Blood Pressure Prevention and Control,
Laboratory Centre for Disease Control at Health Canada, Heart and Stroke Foundation of
Canada. Canadian Medical Association Journal. 1999; 160(9 Suppl):S46-S50.
551 Spence JD, Huff M, Barnett PA. Effects of indapamide versus hydrochlorothiazide on
plasma lipids and lipoproteins in hypertensive patients: a direct comparison. Canadian
Journal of Clinical Pharmacology. 2000; 7(1):32-37.
552 Spinar J, Vitovec J, Soucek M, Dusek L, Pavlik T, Invesigators CORD. CORD:
COmparsion of Recommended Doses of ACE inhibitors and angiotensin II receptor blockers.
Vnitrni Lekarstvi. 2009; 55(5):481-488.
553 Sprafka JM, Strickland D, Gomez-Marin O, Prineas RJ. The effect of cuff size on
blood pressure measurement in adults. Epidemiology. 1991; 2(3):214-217.
554 Staessen JA, Den HE, Celis H, Fagard R, Keary L, G, O'Brien ET. Antihypertensive
treatment based on blood pressure measurement at home or in the physician's office: a
randomized controlled trial. JAMA. 2004; 291(8):955-964.
555 Staessen JA, Fagard R, Thijs L, Celis H, Arabidze GG, Birkenhager WH, Bulpitt CJ,
De Leeuw P, Dollery CT, Fletcher AE, Forette F, Leonetti G, Nachev C, O'Brien ET,
Rosenfeld J, Rodicio JL, Tuomilehto J, Zanchetti A. Randomised double-blind comparison of
placebo and active treatment for older patients with isolated systolic hypertension. Lancet.
1997; 350(9080):757-764.
556 Staessen JA, Thijs L. Development of diagnostic thresholds for automated selfmeasurement of blood pressure in adults. First International Consensus Conference on Blood
Pressure Self-Measurement. Blood Pressure Monitoring. 2000; 5(2):101-109.
557 Staessen JA, Thijs L, Fagard R, O'Brien ET, Clement D, de Leeuw PW, Mancia G,
Nachev C, Palatini P, Parati G, Tuomilehto J, Webster J. Predicting cardiovascular risk using
conventional vs ambulatory blood pressure in older patients with systolic hypertension.
Systolic Hypertension in Europe Trial Investigators. JAMA. 1999; 282(6):539-546.
558 Staessen JA, Wang JG, Thijs L, Celis H, Gasowski J, Fagard RH. Use of
dihydropyridines for antihypertensive treatment in older patients: Evidence from the systolic
hypertension in Europe trial. Journal of Clinical & Basic Cardiology. 2000; 3:15-21.
559 Staffileno BA, Braun LT, Rosenson RS. The accumulative effects of physical activity
in hypertensive post-menopausal women. Journal of Cardiovascular Risk. 2001; 8(5):283290.
560 Stafilas PC, Sarafidis PA, Lasaridis AN, Aletras VH, Niakas D. An economic
evaluation of the 2003 European Society of Hypertension-European Society of Cardiology
guidelines for the management of mild-to-moderate hypertension in Greece. American
Journal of Hypertension. 2005; 18(9):1233-1240.
561 Stamler R, Stamler J, Grimm R, Gosch F, Dyer A, Berman R, Civinelli J, Elmer P,
Fishman J, Van HN, . Trial on control of hypertension by nutritional means: three-year
results. Journal of Hypertension - Supplement. 1984; 2(3):S167-S170.
562 Stenehjem AE, Os I. Reproducibility of blood pressure variability, white-coat effect
and dipping pattern in untreated, uncomplicated and newly diagnosed essential hypertension.
Blood Pressure. 2004; 13(4):214-224.
563 Stergiou GS, Baibas NM, Gantzarou AP, Skeva II, Kalkana CB, Roussias LG,
Mountokalakis TD. Reproducibility of home, ambulatory, and clinic blood pressure:
implications for the design of trials for the assessment of antihypertensive drug efficacy.
American Journal of Hypertension. 2002; 15(2 Pt 1):101-104.
293
Hypertension (partial update)
Reference list
564 Stergiou GS, Baibas NM, Kalogeropoulos PG. Cardiovascular risk prediction based
on home blood pressure measurement: the Didima study. Journal of Hypertension. 2007;
25(8):1590-1596.
565 Stergiou GS, Nasothimiou EG, Kalogeropoulos PG, Pantazis N, Baibas NM. The
optimal home blood pressure monitoring schedule based on the Didima outcome study.
Journal of Human Hypertension. 2010; 24(3):158-164.
566 Stergiou GS, Salgami EV, Tzamouranis DG, Roussias LG. Masked hypertension
assessed by ambulatory blood pressure versus home blood pressure monitoring: is it the same
phenomenon? American Journal of Hypertension. 2005; 18(6):772-778.
567 Stergiou GS, Skeva II, Baibas NM, Kalkana CB, Roussias LG, Mountokalakis TD.
Diagnosis of hypertension using home or ambulatory blood pressure monitoring: comparison
with the conventional strategy based on repeated clinic blood pressure measurements. Journal
of Hypertension. 2000; 18(12):1745-1751.
568 Stergiou GS, Skeva II, Zourbaki AS, Mountokalakis TD. Self-monitoring of blood
pressure at home: how many measurements are needed? Journal of Hypertension. 1998;
16(6):725-731.
569 Stergiou GS, Zourbaki AS, Skeva II, Mountokalakis TD. White coat effect detected
using self-monitoring of blood pressure at home: comparison with ambulatory blood pressure.
American Journal of Hypertension. 1998; 11(7):820-827.
570 Stewart PM. Mineralocorticoid hypertension. Lancet. 1999; 353(9161):1341-1347.
571 Strazzullo P, Siani A, Gugliemi S, Di CA, Galletti F, Cirillo M, Mancini M.
Controlled trial of long-term oral calcium supplementation in essential hypertension.
Hypertension. 1986; 8(11):1084-1088.
572 Streeten DH, Anderson GH, Jr., Howland T, Chiang R, Smulyan H. Effects of thyroid
function on blood pressure. Recognition of hypothyroid hypertension. Hypertension. 1988;
11(1):78-83.
573 Suarez C, Del AC, Garcia-Polo I. Ambulatory blood pressure monitoring: is the
daytime period enough for making clinical decisions? Blood Pressure Monitoring. 2003;
8(6):267-270.
574 Sutcliffe SJ, Fox KF, Wood DA, Sutcliffe A, Stock K, Wright M, Akhras F, Langford
E. Incidence of coronary heart disease in a health authority in London: review of a community
register. British Medical Journal. 2003; 326(7379):20.
575 Suter E, Marti B, Tschopp A, Wanner HU, Wenk C, Gutzwiller F. Effects of selfmonitored jogging on physical fitness, blood pressure and serum lipids: a controlled study in
sedentary middle-aged men. International Journal of Sports Medicine. 1990; 11(6):425-432.
576 Suzuki Y, Kuwajima I, Aono T, Kanemaru A, Nishinaga M, Shibata H, Ozawa T.
Prognostic value of nighttime blood pressure in the elderly: a prospective study of 24-hour
blood pressure. Hypertension Research - Clinical and Experimental. 2000; 23(4):323-330.
577 Svetkey LP, Simons-Morton D, Vollmer WM, Appel LJ, Conlin PR, Ryan DH, Ard J,
Kennedy BM. Effects of dietary patterns on blood pressure: subgroup analysis of the Dietary
Approaches to Stop Hypertension (DASH) randomized clinical trial. Archives of Internal
Medicine. 1999; 159(3):285-293.
578 Svetkey LP, Yarger WE, Feussner JR, DeLong E, Klotman PE. Double-blind,
placebo-controlled trial of potassium chloride in the treatment of mild hypertension.
Hypertension. 1987; 9(5):444-450.
579 Swales JD.Manual of Hypertension. Blackwell Science; 1995
580 Szucs TD, Waeber B, Tomonaga Y. Cost-effectiveness of antihypertensive treatment
in patients 80 years of age or older in Switzerland: an analysis of the HYVET study from a
Swiss perspective. Journal of Human Hypertension. 2010; 24(2):117-123.
294
Hypertension (partial update)
Reference list
581 Takagi Y, Fukase M, Takata S, Fujimi T, Fujita T. Calcium treatment of essential
hypertension in elderly patients evaluated by 24 H monitoring. American Journal of
Hypertension. 1991; 4(10 Pt 1):836-839.
582 Takata Y, Yoshizumi T, Ito Y, Ueno M, Tsukashima A, Iwase M, Kobayashi K,
Fujishima M. Comparison of withdrawing antihypertensive therapy between diuretics and
angiotensin converting enzyme inhibitors in essential hypertensives. American Heart Journal.
1992; 124(6):1574-1580.
583 Tanabe Y, Urata H, Kiyonaga A, Ikeda M, Tanaka H, Shindo M, Arakawa K. Changes
in serum concentrations of taurine and other amino acids in clinical antihypertensive exercise
therapy. Clinical and Experimental Hypertension Part A: Theory and Practice. 1989;
11(1):149-165.
584 Tanji JL, Lew EY, Wong GY, Treguboff C, Ward JA, Amsterdam EA. Dietary
calcium supplementation as a treatment for mild hypertension. Journal of the American Board
of Family Practice. 1991; 4(3):145-150.
585 Taylor AH, Doust J, Webborn N. Randomised controlled trial to examine the effects
of a GP exercise referral programme in Hailsham, East Sussex, on modifiable coronary heart
disease risk factors. Journal of Epidemiology and Community Health. 1998; 52(9):595-601.
586 Taylor CB, Farquhar JW, Nelson E, Agras S. Relaxation therapy and high blood
pressure. Archives of General Psychiatry. 1977; 34(3):339-342.
587 Tedesco MA, Natale F, Calabro R. Effects of monotherapy and combination therapy
on blood pressure control and target organ damage: a randomized prospective intervention
study in a large population of hypertensive patients. Journal of Clinical Hypertension. 2006;
8(9):634-641.
588 Tengs TO, Lin TH. A meta-analysis of quality-of-life estimates for stroke.
Pharmacoeconomics. 2003; 21(3):191-200.
589 The ALLHAT Officers and Co-ordinators for the ALLHAT Collaborative Research
Group. Major cardiovascular events in hypertensive patients randomized to doxazosin vs
chlorthalidone: The antihypertensive and lipid-lowering treatment to prevent heart attack trial
(ALLHAT). JAMA. 2000; 283:1967-1975.
590 The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial
(ALLHAT). Major outcomes in high-risk hypertensive patients randomized to angiotensinconverting enzyme inhibitor or calcium channel blocker vs diuretic. JAMA. 2002;
288(23):2981-2997.
591 The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial
(ALLHAT-LLT). Major outcomes in moderately hypercholesterolemic, hypertensive patients
randomized to Pravastatin vs usual care. JAMA. 2002; 288(23):2998-3007.
592 The CAPPP group. The Captopril Prevention Project: a prospective intervention trial
of angiotensin converting enzyme inhibition in the treatment of hypertension. Journal of
Hypertension. 1990; 8:985-990.
593 The NHS Information Centre Prescribing Support Unit. NHS Prescription Cost
Analysis 2009. The Health and Social Care Information Centre, 2009.
594 The Nordic Diltiazem Study Group. A prospective intervention trials of calcium
antagonist therapy in hypertension. Blood Pressure. 1993; 2(4):312-321.
595 Thijs L, Amery A, Clement D, Cox J, De CP, Fagard R, Fowler G, Guo C, Mancia G,
Marin R, O'Brien E, O'Malley K, Palatini P, Parati G, Petrie J, Ravogli A, Rosenfeld J,
Staessen J, Webster J. Ambulatory blood pressure monitoring in elderly patients with isolated
systolic hypertension. Journal of Hypertension. 1992; 10(7):693-699.
596 Thun MJ, Peto R, Lopez AD, Monaco JH, Henley SJ, Heath CW, Jr., Doll R. Alcohol
consumption and mortality among middle-aged and elderly U.S. adults. New England Journal
of Medicine. 1997; 337(24):1705-1714.
295
Hypertension (partial update)
Reference list
597 Thurm RH, Smith WM. On restting of "Barostats" in hypertensive patients. JAMA.
1967; 201:301-304.
598 Toal CB, Mahon WA, Barnes C, Burelle D. Nifedipine gastrointestinal therapeutic
system (GITS) for hypertensive patients in a primary care setting: results of the Extended
Release Adalat Canadian Trial (EXACT). Clinical Therapeutics. 1997; 19(5):924-935.
599 Torjesen PA, Birkeland KI, Anderssen SA, Hjermann I, Holme I, Urdal P. Lifestyle
changes may reverse development of the insulin resistance syndrome. The Oslo Diet and
Exercise Study: a randomized trial. Diabetes Care. 1997; 20(1):26-31.
600 Trazzi S, Mutti E, Frattola A, Imholz B, Parati G, Mancia G. Reproducibility of noninvasive and intra-arterial blood pressure monitoring: implications for studies on
antihypertensive treatment. Journal of Hypertension. 1991; 9(2):115-119.
601 Tresukosol D, Sriyudhasak O. Amlodipine and hydrochlorothiazide for isolated
systolic hypertension in the Thai elderly. Siriraj Medical Journal. 2005; 57(9):374-379.
602 Trudel X, Brisson C, Larocque B, Milot A. Masked hypertension: different blood
pressure measurement methodology and risk factors in a working population. Journal of
Hypertension. 2009; 27(8):1560-1567.
603 Ungar A, Pepe G, Monami M, Lambertucci L, Torrini M, Baldasseroni S, Tarantini F,
Marchionni N, Masotti G. Isolated ambulatory hypertension is common in outpatients
referred to a hypertension centre. Journal of Human Hypertension. 2004; 18(12):897-903.
604 Urata H, Tanabe Y, Kiyonaga A, Ikeda M, Tanaka H, Shindo M, Arakawa K.
Antihypertensive and volume-depleting effects of mild exercise on essential hypertension.
Hypertension. 1987; 9(3):245-252.
605 Uusitupa M, Louheranta A, Lindstrom J, Valle T, Sundvall J, Eriksson J, Tuomilehto
J. The Finnish Diabetes Prevention Study. British Journal of Nutrition. 2000; 83 Suppl
1:S137-S142.
606 Vaccarino V, Berger AK, Abramson J, Black HR, Setaro JF, Davey JA, Krumholz
HM. Pulse pressure and risk of cardiovascular events in the systolic hypertension in the
elderly program. American Journal of Cardiology. 2001; 88(9):980-986.
607 Van Bortel LM, Bulpitt CJ, Fici F. Quality of life and antihypertensive effect with
nebivolol and losartan. American Journal of Hypertension. 2005; 18(8):1060-1066.
608 van der Steen MS, Lenders JW, Graafsma SJ, den AJ, Thien T. Reproducibility of
ambulatory blood pressure monitoring in daily practice. Journal of Human Hypertension.
1999; 13(5):303-308.
609 van Ittersum FJ, Ijzerman RG, Stehouwer CD, Donker AJ. Analysis of twenty-fourhour ambulatory blood pressure monitoring: what time period to assess blood pressures
during waking and sleeping? Journal of Hypertension. 1995; 13(9):1053-1058.
610 Van Montfrans GA, Karemaker JM, Wieling W, Dunning AJ. Relaxation therapy and
continuous ambulatory blood pressure in mild hypertension: a controlled study. BMJ. 1990;
300(6736):1368-1372.
611 Verberk WJ, Kroon AA, Kessels AG, de Leeuw PW. Home blood pressure
measurement: a systematic review. Journal of the American College of Cardiology. 2005;
46:743-751.
612 Verberk WJ, Kroon AA, Kessels AG, Lenders JW, Thien T, Van Montfrans GA, Smit
AJ, de Leeuw PW. The optimal scheme of self blood pressure measurement as determined
from ambulatory blood pressure recordings. Journal of Hypertension. 2006; 24(8):1541-1548.
613 Verdecchia P, Reboldi G, Porcellati C, Schillaci G, Pede S, Bentivoglio M, Angeli F,
Norgiolini S, Ambrosio G. Risk of cardiovascular disease in relation to achieved office and
ambulatory blood pressure control in treated hypertensive subjects. Journal of the American
College of Cardiology. 2002; 39(5):878-885.
296
Hypertension (partial update)
Reference list
614 Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Pede S, Porcellati C. Ambulatory
pulse pressure: a potent predictor of total cardiovascular risk in hypertension. Hypertension.
1998; 32(6):983-988.
615 Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Porcellati C. Prognostic significance
of the white coat effect. Hypertension. 1997; 29(6):1218-1224.
616 Verdecchia P, Staessen JA, Angeli F, de SG, Achilli A, Ganau A, Mureddu G, Pede S,
Maggioni AP, Lucci D, Reboldi G, Cardio-Sis i. Usual versus tight control of systolic blood
pressure in non-diabetic patients with hypertension (Cardio-Sis): an open-label randomised
trial. Lancet. 2009; 374(9689):525-533.
617 Vogt TM, Appel LJ, Obarzanek E, Moore TJ, Vollmer WM, Svetkey LP, Sacks FM,
Bray GA, Cutler JA, Windhauser MM, Lin PH, Karanja NM. Dietary Approaches to Stop
Hypertension: rationale, design, and methods. DASH Collaborative Research Group. Journal
of the American Dietetic Association. 1999; 99(8 Suppl):S12-S18.
618 Wachtell K, Hornestam B, Lehto M, Slotwiner DJ, Gerdts E, Olsen MH, Aurup P,
Dahlof B, Ibsen H, Julius S, Kjeldsen SE, Lindholm LH, Nieminen MS, Rokkedal J,
Devereux RB. Cardiovascular morbidity and mortality in hypertensive patients with a history
of atrial fibrillation: The Losartan Intervention For End Point Reduction in Hypertension
(LIFE) study. Journal of the American College of Cardiology. 2005; 45(5):705-711.
619 Wachtell K, Lehto M, Gerdts E, Olsen MH, Hornestam B, Dahlof B, Ibsen H, Julius
S, Kjeldsen SE, Lindholm LH, Nieminen MS, Devereux RB. Angiotensin II receptor
blockade reduces new-onset atrial fibrillation and subsequent stroke compared to atenolol: the
Losartan Intervention For End Point Reduction in Hypertension (LIFE) study. Journal of the
American College of Cardiology. 2005; 45(5):712-719.
620 Waeber B, Brunner HR, Metry JM. Compliance with antihypertensive treatment:
implications for practice. Blood Pressure. 1997; 6(6):326-331.
621 Walker AF, Marakis G, Morris AP, Robinson PA. Promising hypotensive effect of
hawthorn extract: a randomized double-blind pilot study of mild, essential hypertension.
Phytotherapy Research. 2002; 16(1):48-54.
622 Wallace JP, Park S, Zakutansky DW, Lehmkuhl LA, Jastremski CA. Time of day to
monitor ambulatory blood pressure affects the outcome. Blood Pressure Monitoring. 2005;
10(1):43-50.
623 Wang JG, Staessen JA, Franklin SS, Fagard R, Gueyffier F. Systolic and diastolic
blood pressure lowering as determinants of cardiovascular outcome. Hypertension. 2005;
45(5):907-913.
624 Wang JG, Staessen JA, Gong L, Liu L. Chinese trial on isolated systolic hypertension
in the elderly. Systolic Hypertension in China (Syst-China) Collaborative Group. Archives of
Internal Medicine. 2000; 160(2):211-220.
625 Ward, S. Statins for the Prevention of Coronary Events. London: National Insititute
for Clinical Excellence, 2005.
626 Ward S, Lloyd JM, Pandor A, Holmes M, Ara R, Ryan A, Yeo W, Payne N. A
systematic review and economic evaluation of statins for the prevention of coronary events.
Health Technology Assessment. 2007; 11(14):1-iv.
627 Weinberger MH, Wagner UL, Fineberg NS. The blood pressure effects of calcium
supplementation in humans of known sodium responsiveness. American Journal of
Hypertension. 1993; 6(9):799-805.
628 Weir MR. Major outcomes in high risk hypertensive patients randomized to
angiotensin-converting enzyme inhibitor or CCB vs diuretic: the Antihypertensive and Lipidlowering Treatment to Prevent Heart Attack Trial (ALLHAT). Current Hypertension Reports.
2003; 5(5):405-407.
297
Hypertension (partial update)
Reference list
629 Weitzman D, Goldbourt U. The significance of various blood pressure indices for
long-term stroke, coronary heart disease, and all-cause mortality in men: the Israeli Ischemic
Heart Disease study. Stroke. 2006; 37(2):358-363.
630 West R. Assessment of evidence versus consensus or prejudice. Journal of
Epidemiology and Community Health. 1992; 46(4):321-322.
631 Whelton PK, Appel LJ, Espeland MA, Applegate WB, Ettinger WH, Jr., Kostis JB,
Kumanyika S, Lacy CR, Johnson KC, Folmar S, Cutler JA. Sodium reduction and weight loss
in the treatment of hypertension in older persons: a randomized controlled trial of
nonpharmacologic interventions in the elderly (TONE). TONE Collaborative Research
Group. JAMA. 1998; 279(11):839-846.
632 Whelton PK, He J. Potassium in preventing and treating high blood pressure.
Seminars in Nephrology. 1999; 19(5):494-499.
633 Whelton PK, He J, Cutler JA, Brancati FL, Appel LJ, Follmann D, Klag MJ. Effects
of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials.
JAMA. 1997; 277(20):1624-1632.
634 Whelton SP, Chin A, Xin X, He J. Effect of aerobic exercise on blood pressure: a
meta-analysis of randomized, controlled trials. Annals of Internal Medicine. 2002;
136(7):493-503.
635 White WB. Ambulatory blood pressure as a predictor of target organ disease and
outcome in the hypertensive patient. Blood Pressure Monitoring. 1999; 4(3-4):181-184.
636 White WB, Dey HM, Schulman P. Assessment of the daily blood pressure load as a
determinant of cardiac function in patients with mild-to-moderate hypertension. American
Heart Journal. 1989; 118(4):782-795.
637 Whitehead A. Meta-Analysis of Controlled Clinical Trials. Chichester, UK: Wiley,
2002: 192.
638 Whitty CJ, Brunner EJ, Shipley MJ, Hemingway H, Marmot MG. Differences in
biological risk factors for cardiovascular disease between three ethnic groups in the Whitehall
II study. Atherosclerosis. 1999; 142(2):279-286.
639 Widimsky J. Treatment of very elderly hypertensives significantly reduces total
mortality and the risk of death from stroke. Results of the HYVET trial. Cor Et Vasa. 2008;
50(9):354-357.
640 Wikstrand J, Warnold I, Olsson G, Tuomilehto J, Elmfeldt D, Berglund G. Primary
prevention with Metoprolol in patients with hypertension: mortality results from the MAPHY
study. JAMA. 1988; 259(13):1976-1982.
641 Wilhelmsen L, Berglund G, Elmfeldt D, Fitzsimons T, Holzgreve H, Hosie J,
Hörnkvist PE, Pennert K, Tuomilehto J, Wedel H. Beta Blockers versus diuretics in
hypertensive men: main results from the HAPPHY Trial. Journal of Hypertension. 1987;
5(5):561-572.
642 Williams B, Poulter NR, Brown MJ, Davis M, McInnes GT, Potter JF, Sever PS,
Thom SM. Guidelines for management of hypertension: report of the fourth working party of
the British Hypertension Society, 2004-BHS IV. Journal of Human Hypertension. 2004;
18(3):139-185.
643 Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB.
Prediction of coronary heart disease using risk factor categories. Circulation. 1998;
97(18):1837-1847.
644 Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL, Johnston CI,
McNeil JJ, Macdonald GJ, Marley JE, Morgan TO, West MJ. A comparison of outcomes with
angiotensin-converting--enzyme inhibitors and diuretics for hypertension in the elderly. New
England Journal of Medicine. 2003; 348(7):583-592.
298
Hypertension (partial update)
Reference list
645 Wirell MP, Wester PO, Stegmayr BG. Nutritional dose of magnesium in hypertensive
patients on beta blockers lowers systolic blood pressure: a double-blind, cross-over study.
Journal of Internal Medicine. 1994; 236(2):189-195.
646 Witteman JC, Grobbee DE, Derkx FH, Bouillon R, de Bruijn AM, Hofman A.
Reduction of blood pressure with oral magnesium supplementation in women with mild to
moderate hypertension. American Journal of Clinical Nutrition. 1994; 60(1):129-135.
647 World Health Organisation. Adherence to Long Term Therapies: Evidence for Action.
WHO, 2003.
648 Wynder EL, Graham EA. Tobacco smoking as a possible etiologic factor in
bronchiogenic carcinoma; a study of 684 proved cases. JAMA. 1950; 143(4):329-336.
649 Youman P, Wilson K, Harraf F, Kalra L. The economic burden of stroke in the United
Kingdom. Pharmacoeconomics. 2003; 21 Suppl 1:43-50.
650 Yui Y, Sumiyoshi T, Kodama K, Hirayama A, Nonogi H, Kanmatsuse K, Origasa H,
Iimura O, Ishii M, Saruta T, Arakawa K, Hosoda S, Kawai C. Nifedipine retard was as
effective as angiotensin converting enzyme inhibitors in preventing cardiac events in highrisk hypertensive patients with diabetes and coronary artery disease: the Japan Multicenter
Investigation for Cardiovascular Diseases-B (JMIC-B) subgroup analysis. Hypertension
Research - Clinical and Experimental. 2004; 27(7):449-456.
651 Yui Y, Sumiyoshi T, Kodama K, Hirayama A, Nonogi H, Kanmatsuse K, Origasa H,
Iimura O, Ishii M, Saruta T, Arakawa K, Hosoda S, Kawai C, Japan Multicenter Investigation
for Cardiovascular Diseases-. Comparison of nifedipine retard with angiotensin converting
enzyme inhibitors in Japanese hypertensive patients with coronary artery disease: the Japan
Multicenter Investigation for Cardiovascular Diseases-B (JMIC-B) randomized trial.
Hypertension Research - Clinical and Experimental. 2004; 27(3):181-191.
652 Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an
angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk
patients. The Heart Outcomes Prevention Evaluation Study Investigators. New England
Journal of Medicine. 2000; 342(3):145-153.
653 Yusuf S, Teo K, Pogue J, Dyal L, Copland I, Schumacher H, Dagenais G, Sleight P,
Anderson C. Telmisartan, ramipril, or both in patients at high risk for vascular events. New
England Journal of Medicine. 2008; 358(15):1547-1559.
654 Zakopoulos NA, Nanas SN, Lekakis JP, Vemmos KN, Kotsis VT, Pitiriga VC,
Stamatelopoulos SF, Moulopoulos SD. Reproducibility of ambulatory blood pressure
measurements in essential hypertension. Blood Pressure Monitoring. 2001; 6(1):41-45.
655 Zanchetti A. Bottom blood pressure or bottom cardiovascular risk? how far can
cardiovascular risk be reduced? Journal of Hypertension. 2009; 27(8):1509-1520.
656 Zanchetti A, Bond MG, Hennig M, Neiss A, Mancia G, Dal Palu C, Hansson L,
Magnani B, Rahn KH, Reid JL, Rodicio J, Safar M, Eckes L, Rizzini P, European Lacidipine
Study on Atherosclerosis investigators. Calcium antagonist lacidipine slows down progression
of asymptomatic carotid atherosclerosis: principal results of the European Lacidipine Study
on Atherosclerosis (ELSA), a randomized, double-blind, long-term trial. Circulation. 2002;
106(19):2422-2427.
657 Zanchetti A, Crepaldi G, Bond MG, Gallus G, Veglia F, Mancia G, Ventura A,
Baggio G, Sampieri L, Rubba P, Sperti G, Magni A, PHYLLIS I. Different effects of
antihypertensive regimens based on fosinopril or hydrochlorothiazide with or without lipid
lowering by pravastatin on progression of asymptomatic carotid atherosclerosis: principal
results of PHYLLIS--a randomized double-blind trial. Stroke. 2004; 35(12):2807-2812.
658 Zanchetti A, Rosei EA, et al. The Verapamil in Hypertension and Atherosclerosis
Study (VHAS): results of long-term randomized treatment with either verapamil or
chlorthalidone on carotid intima-media thickness. Journal of Hypertension. 1998; 16:16671676.
299
Hypertension (partial update)
Glossary
659 Zemel PC, Zemel MB, Urberg M, Douglas FL, Geiser R, Sowers JR. Metabolic and
hemodynamic effects of magnesium supplementation in patients with essential hypertension.
American Journal of Clinical Nutrition. 1990; 51(4):665-669.
660 Zoccali C, Mallamaci F, Delfino D, Ciccarelli M, Parlongo S, Iellamo D, Moscato D,
Maggiore Q. Double-blind randomized, crossover trial of calcium supplementation in
essential hypertension. Journal of Hypertension. 1988; 6(6):451-455.
661 Zurawski RM, Smith TW, Houston BK. Stress management for essential
hypertension: comparison with a minimally effective treatment, predictors of response to
treatment, and effects on reactivity. Journal of Psychosomatic Research. 1987; 31(4):453462.
662 Zyczynski TM, Leidy NK, Kong BW, Helaszek CT, Michelson EL. Effects of
candesartan cilexetil on health-related quality of life in black patients with systemic
hypertension in the ABC Trial. Heart Disease. 2000; 2(6):400-406.
Glossary
Term
Definition
Ambulatory blood pressure
monitoring (ABPM)
A technique for measuring BP while an individual goes about their normal
daily activities
Abstract
Summary of a study, which may be published alone or as an introduction to a
full scientific paper.
Aerobic exercise
Exercise requiring increased oxygen
Algorithm (in guidelines)
A flow chart of the clinical decision pathway described in the guideline, where
decision points are represented with boxes, linked with arrows.
Allocation concealment
The process used to prevent advance knowledge of group assignment in a RCT.
The allocation process should be impervious to any influence by the individual
making the allocation, by being administered by someone who is not
responsible for recruiting participants.
Angina pectoris:
A strangling pain in the chest due to reduced blood flowing to the heart
muscles
Antihypertensive
Drug used to lower blood pressure
Applicability
The degree to which the results of an observation, study or review are likely to
hold true in a particular clinical practice setting.
Arm (of a clinical study)
Sub-section of individuals within a study who receive one particular
intervention, for example placebo arm
Arrhythmia
A variation in the normal rhythm of the heart
Association
Statistical relationship between two or more events, characteristics or other
variables. The relationship may or may not be causal.
Auscultation
Examination of the internal organs by listening to the sound produced
Baseline
The initial set of measurements at the beginning of a study (after run-in period
where applicable), with which subsequent results are compared.
Before-and-after study
A study that investigates the effects of an intervention by measuring particular
characteristics of a population both before and after taking the intervention, and
assessing any change that occurs.
Bias
Systematic (as opposed to random) deviation of the results of a study from the
‘true’ results that is caused by the way the study is designed or conducted.
Biofeedback
Sight or sound information letting the individual know how an aspect of their
body is functioning
Blinding
Keeping the study participants, caregivers, researchers and outcome assessors
unaware about the interventions to which the participants have been allocated
in a study.
Blood pressure
Force exerted by blood against the walls of blood vessels
Caffeine
A substance which acts as a stimulant, found in coffee and tea
300
Hypertension (partial update)
Glossary
Term
Definition
Calcium
An element necessary for normal body function; most of our calcium intake
comes from milk and milk products
Calorie
A unit of heat, used as a measure of energy supplied by food
Cardiovascular Disease
Disease affecting the heart or blood vessels
Carer (caregiver)
Someone other than a health professional who is involved in caring for a
person with a medical condition.
Case-control study
Comparative observational study in which the investigator selects individuals
who have experienced an event (For example, developed a disease) and others
who have not (controls), and then collects data to determine previous exposure
to a possible cause.
Case-series
Report of a number of cases of a given disease, usually covering the course of
the disease and the response to treatment. There is no comparison (control)
group of patients.
Cerebrovascular accident
Stroke (part of the brain is damaged due to lack of oxygen)
Cerebrovascular disease
Narrowing of the arteries supplying blood to the brain
Clinical efficacy
The extent to which an intervention is active when studied under controlled
research conditions.
Clinical effectiveness
The extent to which an intervention produces an overall health benefit in
routine clinical practice.
Clinician
A healthcare professional providing direct patient care, for example doctor,
nurse or physiotherapist.
Cochrane Review
The Cochrane Library consists of a regularly updated collection of evidencebased medicine databases including the Cochrane Database of Systematic
Reviews (reviews of randomised controlled trials prepared by the Cochrane
Collaboration).
Cognitive
Describing mental processes
Cohort study
A retrospective or prospective follow-up study. Groups of individuals to be
followed up are defined on the basis of presence or absence of exposure to a
suspected risk factor or intervention. A cohort study can be comparative, in
which case two or more groups are selected on the basis of differences in their
exposure to the agent of interest.
Comorbidity
Co-existence of more than one disease or an additional disease (other than that
being studied or treated) in an individual.
Comparability
Similarity of the groups in characteristics likely to affect the study results (such
as health status or age).
Concordance
This is a recent term whose meaning has changed. It was initially applied to the
consultation process in which doctor and patient agree therapeutic decisions
that incorporate their respective views, but now includes patient support in
medicine taking as well as prescribing communication. Concordance reflects
social values but does not address medicine-taking and may not lead to
improved adherence.
Confidence interval (CI)
A range of values for an unknown population parameter with a stated
‘confidence’ (conventionally 95%) that it contains the true value. The interval
is calculated from sample data, and generally straddles the sample estimate.
The ‘confidence’ value means that if the method used to calculate the interval
is repeated many times, then that proportion of intervals will actually contain
the true value.
Confounding
In a study, confounding occurs when the effect of an intervention on an
outcome is distorted as a result of an association between the population or
intervention or outcome and another factor (the ‘confounding variable’) that
can influence the outcome independently of the intervention under study.
Consensus methods
Techniques that aim to reach an agreement on a particular issue. Consensus
methods may used when there is a lack of strong evidence on a particular topic.
301
Hypertension (partial update)
Glossary
Term
Definition
Control group
A group of patients recruited into a study that receives no treatment, a
treatment of known effect, or a placebo (dummy treatment) - in order to
provide a comparison for a group receiving an experimental treatment, such as
a new drug.
Coronary heart disease
Heart disease due to narrowing of the arteries which provide the heart's blood
supply; may manifest as angina or heart attack
Cost benefit analysis
A type of economic evaluation where both costs and benefits of healthcare
treatment are measured in the same monetary units. If benefits exceed costs, the
evaluation would recommend providing the treatment.
Cost-consequences analysis
(CCA)
A type of economic evaluation where various health outcomes are reported in
addition to cost for each intervention, but there is no overall measure of health
gain.
Cost-effectiveness analysis
(CEA)
An economic study design in which consequences of different interventions are
measured using a single outcome, usually in ‘natural’ units (For example, lifeyears gained, deaths avoided, heart attacks avoided, cases detected). Alternative
interventions are then compared in terms of cost per unit of effectiveness.
Cost-effectiveness model
An explicit mathematical framework, which is used to represent clinical
decision problems and incorporate evidence from a variety of sources in order
to estimate the costs and health outcomes.
Cost-utility analysis (CUA)
A form of cost-effectiveness analysis in which the units of effectiveness are
quality-adjusted life-years (QALYs).
Credible Interval
The Bayesian equivalent of a confidence interval.
Decision analysis
An explicit quantitative approach to decision making under uncertainty, based
on evidence from research. This evidence is translated into probabilities, and
then into diagrams or decision trees which direct the clinician through a
succession of possible scenarios, actions and outcomes.
Diastolic blood pressure
The lowest blood pressure during each heartbeat (e.g. 80 if blood pressure is
140/80 mmHg)
Discounting
Costs and perhaps benefits incurred today have a higher value than costs and
benefits occurring in the future. Discounting health benefits reflects individual
preference for benefits to be experienced in the present rather than the future.
Discounting costs reflects individual preference for costs to be experienced in
the future rather than the present.
Dominance
An intervention is said to be dominated if there is an alternative intervention
that is both less costly and more effective.
Dose titration
Change in the dose of a drug
Drop-out
A participant who withdraws from a trial before the end.
Economic evaluation
Comparative analysis of alternative health strategies (interventions or
programmes) in terms of both their costs and consequences.
Effect (as in effect
measure, treatment effect,
estimate of effect, effect
size)
The observed association between interventions and outcomes or a statistic to
summarise the strength of the observed association.
Effectiveness
See ‘Clinical effectiveness’.
Efficacy
See ‘Clinical efficacy’.
Epidemiological study
The study of a disease within a population, defining its incidence and
prevalence and examining the roles of external influences (For example,
infection, diet) and interventions.
EQ-5D (EuroQol-5D)
A standardise instrument used to measure a health outcome. It provides a single
index value for health status.
Essential hypertension
High blood pressure which is not due to a known underlying disease
Excessive alcohol
consumption
Over 21 units/week for men; over 14 units/week for women
302
Hypertension (partial update)
Glossary
Term
Definition
Excessive coffee
consumption
Over 5 cups/day
Evidence
Information on which a decision or guidance is based. Evidence is obtained
from a range of sources including randomised controlled trials, observational
studies, expert opinion (of clinical professionals and/or patients).
Exclusion criteria
(literature review)
Explicit standards used to decide which studies should be excluded from
consideration as potential sources of evidence.
Exclusion criteria (clinical
study)
Criteria that define who is not eligible to participate in a clinical study.
Extended dominance
If Option A is both more clinically effective than Option B and has a lower cost
per unit of effect, when both are compared with a do-nothing alternative then
Option A is said to have extended dominance over Option B. Option A is
therefore more efficient and should be preferred, other things remaining equal.
Extrapolation
In data analysis, predicting the value of a parameter outside the range of
observed values.
Follow-up
Observation over a period of time of an individual, group or initially defined
population whose appropriate characteristics have been assessed in order to
observe changes in health status or health-related variables.
Generalisability
The extent to which the results of a study based on measurement in a particular
patient population and/or a specific context hold true for another population
and/or in a different context. In this instance, this is the degree to which the
guideline recommendation is applicable across both geographical and
contextual settings. For instance, guidelines that suggest substituting one form
of labour for another should acknowledge that these costs might vary across the
country.
Gold standard
See
‘Reference standard’.
GRADE / GRADE profile A system developed by the GRADE Working Group
to address the shortcomings of present grading systems in healthcare. The
GRADE system uses a common, sensible and transparent approach to grading
the quality of evidence. The results of applying the GRADE system to clinical
trial data are displayed in a table known as a GRADE profile.
Harms
Adverse effects of an intervention.
Health economics
The study of the allocation of scarce resources among alternative healthcare
treatments. Health economists are concerned with both increasing the average
level of health in the population and improving the distribution of health.
Health-related quality of
life (HRQoL)
A combination of an individual’s physical, mental and social well-being; not
merely the absence of disease.
Heart failure
Reduction in the heart's pumping efficiency, leading to accumulation of fluid in
the lungs and body, causing fatigue, breathlessness and leg swelling
Heterogeneity
Or lack
of homogeneity.
The term is used in meta-analyses and systematic reviews when the results or
estimates of effects of treatment from separate studies seem to be very different
– in terms of the size of treatment effects or even to the extent that some
indicate beneficial and others suggest adverse treatment effects. Such results
may occur as a result of differences between studies in terms of the patient
populations, outcome measures, definition of variables or duration of followup.
Hypertension
High blood pressure
Imprecision
Results are imprecise when studies include relatively few patients and few
events and thus have wide confidence intervals around the estimate of effect.
Inclusion criteria (literature
review)
Explicit criteria used to decide which studies should be considered as potential
sources of evidence.
Incremental analysis
The analysis of additional costs and additional clinical outcomes with different
interventions.
Incremental cost
The mean cost per patient associated with an intervention minus the mean cost
per patient associated with a comparator intervention.
303
Hypertension (partial update)
Glossary
Term
Definition
Incremental costeffectiveness ratio (ICER)
The difference in the mean costs in the population of interest divided by the
differences in the mean outcomes in the population of interest for one treatment
compared with another.
Incremental net benefit
(INB)
The value (usually in monetary terms) of an intervention net of its cost
compared with a comparator intervention. The INB can be calculated for a
given cost-effectiveness (willingness to pay) threshold. If the threshold is
£20,000 per QALY gained then the INB is calculated as: (£20,000 x QALYs
gained) – Incremental cost.
Indirectness
The available evidence is different to the review question being addressed, in
terms of PICO (population, intervention, comparison and outcome).
Intention to treat analysis
(ITT)
A strategy for analysing data from a randomised controlled trial. All
participants are included in the arm to which they were allocated, whether or
not they received (or completed) the intervention given to that arm. Intentionto-treat analysis prevents bias caused by the loss of participants, which may
disrupt the baseline equivalence established by randomisation and which may
reflect non-adherence to the protocol.
Intervention
Healthcare action intended to benefit the patient, for example, drug treatment,
surgical procedure, psychological therapy.
Intraoperative
The period of time during a surgical procedure.
Ischaemic heart disease
See Coronary heart disease
Kappa statistic
A statistical measure of inter-rater agreement that takes into account the
agreement occurring by chance.
Length of stay
The total number of days a participant stays in hospital.
Licence
See ‘Product licence’.
Lifestyle intervention
A measure to change a participant's behaviour in order to improve their health
(e.g. exercise to reduce blood pressure)
Life-years gained
Mean average years of life gained per person as a result of the intervention
compared with an alternative intervention.
Likelihood ratio
The likelihood ratio combines information about the sensitivity and specificity.
It tells you how much a positive or negative result changes the likelihood that a
patient would have the disease. The likelihood ratio of a positive test result
(LR+) is sensitivity divided by 1- specificity.
Lipid lowering drugs
Drugs used to lower the level of fats in the blood
Long-term care
Loss to follow-up
Residential care in a home that may include skilled nursing care and help with
everyday activities. This includes nursing homes and residential homes.
The loss of participants during the course of a study.
Magnesium
An element necessary for normal body function; found in food
Markov model
A method for estimating long-term costs and effects for recurrent or chronic
conditions, based on health states and the probability of transition between
them within a given time period (cycle).
Meta-analysis
A statistical technique for combining (pooling) the results of a number of
studies that address the same question and report on the same outcomes to
produce a summary result. The aim is to derive more precise and clear
information from a large data pool. It is generally more reliably likely to
confirm or refute a hypothesis than the individual trials.
Monotherapy
Use of only one drug (rather than two or more)
Multivariate model
A statistical model for analysis of the relationship between two or more
predictor (independent) variables and the outcome (dependent) variable.
Negative predictive value
(NPV) [In
screening/diagnostic tests:]
A measure of the usefulness of a screening/diagnostic test. It is the proportion
of those with a negative test result who do not have the disease, and can be
interpreted as the probability that a negative test result is correct.
Normotension
Blood pressure that is within the normal range
304
Hypertension (partial update)
Glossary
Term
Definition
Number needed to treat
(NNT)
The number of patients that who on average must be treated to prevent a single
occurrence of the outcome of interest.
Observational study
Retrospective or prospective study in which the investigator observes the
natural course of events with or without control groups; for example, cohort
studies and case–control studies.
Odds ratio
A measure of treatment effectiveness. The odds of an event happening in the
treatment group, expressed as a proportion of the odds of it happening in the
control group. The 'odds' is the ratio of events to non-events.
Opportunity cost
The loss of other health care programmes displaced by investment in or
introduction of another intervention. This may be best measured by the health
benefits that could have been achieved had the money been spent on the next
best alternative healthcare intervention.
Oscilllometry
The measurement of blood pressure using an electronic device rather than by
listening to Korotkoff sounds (auscultation)
Outcome
Measure of the possible results that may stem from exposure to a preventive or
therapeutic intervention. Outcome measures may be intermediate endpoints or
they can be final endpoints. See ‘Intermediate outcome’.
P-value
The probability that an observed difference could have occurred by chance,
assuming that there is in fact no underlying difference between the means of
the observations. If the probability is less than 1 in 20, the P value is less than
0.05; a result with a P value of less than 0.05 is conventionally considered to be
‘statistically significant’.
Perioperative
The period from admission through surgery until discharge, encompassing the
pre-operative and post-operative periods.
Peripheral vascular disease
Narrowing of the arteries providing circulation to the legs
Placebo
An inactive and physically identical medication or procedure used as a
comparator in controlled clinical trials.
Polypharmacy
The use or prescription of multiple medications.
Positive predictive value
(PPV)
In screening/diagnostic tests: A measure of the usefulness of a
screening/diagnostic test. It is the proportion of those with a positive test result
who have the disease, and can be interpreted as the probability that a positive
test result is correct.
Postoperative
Pertaining to the period after patients leave the operating theatre, following
surgery.
Post-test probability
For diagnostic tests. The proportion of patients with that particular test result
who have the target disorder.
Potassium
An element necessary for normal body function; found in food
Power (statistical)
The ability to demonstrate an association when one exists. Power is related to
sample size; the larger the sample size, the greater the power and the lower the
risk that a possible association could be missed.
Preoperative
The period before surgery commences.
Pre-test probability
For diagnostic tests. The proportion of people with the target disorder in the
population at risk at a specific time point or time interval. Prevalence may
depend on how a disorder is diagnosed.
Primary care
Healthcare delivered to patients outside hospitals. Primary care covers a range
of services provided by general practitioners, nurses, dentists, pharmacists,
opticians and other healthcare professionals.
Primary outcome
The outcome of greatest importance, usually the one in a study that the power
calculation is based on.
Product licence
An authorisation from the MHRA to market a medicinal product.
Prognosis
A probable course or outcome of a disease. Prognostic factors are patient or
disease characteristics that influence the course. Good prognosis is associated
with low rate of undesirable outcomes; poor prognosis is associated with a high
305
Hypertension (partial update)
Glossary
Term
Definition
rate of undesirable outcomes.
Prospective study
A study in which people are entered into the research and then followed up
over a period of time with future events recorded as they happen. This contrasts
with studies that are retrospective.
Publication bias
Also known as reporting bias. A bias caused by only a subset of all the relevant
data being available. The publication of research can depend on the nature and
direction of the study results. Studies in which an intervention is not found to
be effective are sometimes not published. Because of this, systematic reviews
that fail to include unpublished studies may overestimate the true effect of an
intervention. In addition, a published report might present a biased set of results
(e.g. only outcomes or sub-groups where a statistically significant difference
was found.
Quality of life
See ‘Health-related quality of life’.
Quality-adjusted life year
(QALY)
An index of survival that is adjusted to account for the patient’s quality of life
during this time. QALYs have the advantage of incorporating changes in both
quantity (longevity/mortality) and quality (morbidity, psychological,
functional, social and other factors) of life. Used to measure benefits in costutility analysis. The QALYs gained are the mean QALYs associated with one
treatment minus the mean QALYs associated with an alternative treatment.
Quick Reference Guide
An abridged version of NICE guidance, which presents the key priorities for
implementation and summarises the recommendations for the core clinical
audience.
Randomisation
Allocation of participants in a research study to two or more alternative groups
using a chance procedure, such as computer-generated random numbers. This
approach is used in an attempt to ensure there is an even distribution of
participants with different characteristics between groups and thus reduce
sources of bias.
Randomised controlled trial
(RCT)
A comparative study in which participants are randomly allocated to
intervention and control groups and followed up to examine differences in
outcomes between the groups.
Rapid atrial fibrillation
A rapid irregular heartbeat
RCT
See ‘Randomised controlled trial’.
Receiver operated
characteristic (ROC) curve
A graphical method of assessing the accuracy of a diagnostic test. Sensitivity Is
plotted against 1-specificity. A perfect test will have a positive, vertical linear
slope starting at the origin. A good test will be somewhere close to this ideal.
Reference standard
The test that is considered to be the best available method to establish the
presence or absence of the outcome – this may not be the one that is routinely
used in practice.
Relative risk (RR)
The number of times more likely or less likely an event is to happen in one
group compared with another (calculated as the risk of the event in group A/the
risk of the event in group B).
Renin-Angiotensin System
Renin is an enzyme produced by the kidney and has an important role in
hypertension. Renin converts a protein in the blood called angiotensinogen into
angiotensin I. This is then turned into angiotensin II by angiotensin converting
enzyme in the lungs. Angiotensin II reduces the size of the blood vessels
(increasing blood pressure) and triggers the release of a hormone called
aldosterone. Aldosterone is responsible for the retention of water and salt
(which further increase blood pressure).
Reporting bias
See publication bias.
Resistant hypertension
Someone whose blood pressure is not controlled to <140/90mmHg, despite
optimal or best tolerated doses of third line treatment
Resource implication
The likely impact in terms of finance, workforce or other NHS resources.
Retrospective study
A retrospective study deals with the present/ past and does not involve studying
future events. This contrasts with studies that are prospective.
306
Hypertension (partial update)
Glossary
Term
Definition
Review question
In guideline development, this term refers to the questions about treatment and
care that are formulated to guide the development of evidence-based
recommendations.
Secondary outcome
An outcome used to evaluate additional effects of the intervention deemed a
priori as being less important than the primary outcomes.
Selection bias
A systematic bias in selecting participants for study groups, so that the groups
have differences in prognosis and/or therapeutic sensitivities at baseline.
Randomisation (with concealed allocation) of patients protects against this bias.
Sensitivity
Sensitivity or recall rate is the proportion of true positives which are correctly
identified as such. For example in diagnostic testing it is the proportion of true
cases that the test detects.
See the related term ‘Specificity’
Sensitivity analysis
A means of representing uncertainty in the results of economic evaluations.
Uncertainty may arise from missing data, imprecise estimates or
methodological controversy. Sensitivity analysis also allows for exploring the
generalisability of results to other settings. The analysis is repeated using
different assumptions to examine the effect on the results.
One-way simple sensitivity analysis (univariate analysis): each parameter is
varied individually in order to isolate the consequences of each parameter on
the results of the study.
Multi-way simple sensitivity analysis (scenario analysis): two or more
parameters are varied at the same time and the overall effect on the results is
evaluated.
Threshold sensitivity analysis: the critical value of parameters above or below
which the conclusions of the study will change are identified.
Probabilistic sensitivity analysis: probability distributions are assigned to the
uncertain parameters and are incorporated into evaluation models based on
decision analytical techniques (For example, Monte Carlo simulation).
Significance (statistical)
A result is deemed statistically significant if the probability of the result
occurring by chance is less than 1 in 20 (p <0.05).
Specificity
The proportion of true negatives that a correctly identified as such. For
example in diagnostic testing the specificity is the proportion of non-cases
incorrectly diagnosed as cases.
See related term ‘Sensitivity’.
In terms of literature searching a highly specific search is generally narrow and
aimed at picking up the key papers in a field and avoiding a wide range of
papers.
Sphygmomanometer
A device used to measure blood pressure
Stakeholder
Those with an interest in the use of the guideline. Stakeholders include
manufacturers, sponsors, healthcare professionals, and patient and carer groups.
Stepped care
A drug intervention where the dose of the drugs can be increased and/or other
drugs could be added
Systematic review
Research that summarises the evidence on a clearly formulated question
according to a pre-defined protocol using systematic and explicit methods to
identify, select and appraise relevant studies, and to extract, collate and report
their findings. It may or may not use statistical meta-analysis.
Systolic blood pressure
The peak blood pressure during each heartbeat (e.g. 140 if blood pressure is
140/80 mmHg)
Time horizon
The time span over which costs and health outcomes are considered in a
decision analysis or economic evaluation.
Toxicity
The unwanted side-effects of drug treatment. These may vary from mild and/or
self-limiting through to chronic and/or severe. Drugs are studied extensively
before use in patients to understand (and avoid) the circumstances when they
may become inappropriately toxic to patients.
307
Hypertension (partial update)
Glossary
Term
Definition
Transient ischaemic attack
Temporary paralysis, numbness, speech difficulty or other neurological
symptoms that start suddenly and recover within 24 hours
Treatment allocation
Assigning a participant to a particular arm of the trial.
Univariate
Analysis which separately explores each variable in a data set.
Utility
A measure of the strength of an individual’s preference for a specific health
state in relation to alternative health states. The utility scale assigns numerical
values on a scale from 0 (death) to 1 (optimal or ‘perfect’ health). Health states
can be considered worse than death and thus have a negative value.
Withdrawal
Failure or refusal to take the assigned treatment (e.g. because of side effects or
dislike of treatment)
308
Hypertension (partial update)
Scope
Appendices
Appendix A: Scope
309
Hypertension (partial update)
Scope
310
Hypertension (partial update)
Scope
311
Hypertension (partial update)
Scope
312
Hypertension (partial update)
Scope
313
Hypertension (partial update)
Scope
314
Hypertension (partial update)
Scope
315
Hypertension (partial update)
Scope
316
Hypertension (partial update)
Declarations of Interest
Appendix B: Declarations of Interest
GDG meeting
Declaration
Action
 None made.
None required.
Bryan Williams
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
Declared 13th August 2009
 I have received travel expenses and speaker
honoraria for invited lectures, workshops and
small group meetings on hypertension at national
and international conferences. ln most cases the
funds for this (as for most congresses) are via the
pharmaceutical industry, although usually paid via
the conference agency. The travel and subsistence
costs are in accordance with the ABPI code. ln this
context, the hypertension guideline does not deal
with specific products unlike specific appraisals,
instead, it deals with classes of drug treatments all
of which are now generically available, or will be
by the time the guideline deliberates.
 My University and hospital trust has received
investigator initiated research grants from
pharmaceuticals, usually in the context of my
participation as a member of the scientific steering
committee of large-scale clinical trials in
hypertension and NIHR adopted studies which
provide the evidence base for guideline
development. The funds follow the usual
institutional and NIHR formulae to meet the costs
and overheads of conducting these studies and
related meetings. ln the past year, these have been
from Servier,Pfizer and Novartis. My department
also receives funding from government bodies
such as MRC and NIHR on the same basis.
 I am a past-president and member of the British
Hypertension Society. I am a member of the
European and lnternational Societies of
Hypertension. I am founding Trustee (pro-bono) of
the Patient's charity and lnformation service; the
Blood Pressure Association (BPA). I have lead
authored previous national guidelines from the
British Hyperlension Society (BMJ 2004) and was
previously clinical advisor to the NICE
hypertension GDG in 2004 and 2006. I have been
the author of many reports from major clinical
trials in hypertension.
Helen Williams
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
 I have received speaker honoraria or attended
advisory boards for the following companies:
Boehringer-Ingelheim, Pfizer, Servier, SanofiAventis, Solvay, Takeda, but none during the
development of the guideline that are related to
hypertension.
 Non-specific funding has been received to support
a project to improve prescribing in cardiac
rehabilitation services for South London Cardiac
and Stroke Network from: Boehringer-Ingelheim,
317
None required.
Hypertension (partial update)
Declarations of Interest
GDG meeting
Declaration
Pfizer, Servier, Sanofi-Aventis, Solvay
Declared on 20th January, 2011
 April 2010: received an honorarium from Bayer
for participation in an expert panel meeting for a
drug unrelated to hypertension
 June 2010: Received an honorarium from Sanofi
Aventis for presenting at a meeting focusing on the
treatment of atrial fibrillation
 July and Nov 2010: Attended and presented at an
expert panel meeting supported by Servier
focusing on the management on angina
 Sept 2010: presented at the Primary Care
Cardiovascular Society at a satellite meeting
focusing on Acute Coronary Syndromes supported
by Lilly / Daiichi Sankyo
 Nov 2010: Received an honorarium payment from
Sanofi-Aventis for providing training on the
management of atrial fibrillation
Action
Jane Northedge
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
 None made.
None required.
 None made.
None required.
 Clinical Directorship paid by London Genetics
2009-2010. This is a company designed to harness
genetic research talents across London to enable
new drug discovery and prediction of medicine
response- but with no direct interest in any drug
therapies and in particular no interests in licensed
or generic antihypertensives currently available. I
represent seven London Universities on how to
draw genetic research opportunities into London.
Renumeration is up to £18,000 per year based on
work done.
 My clinical trials unit has undertaken the
following study on anti-hypertensives: The
Accelerate Trial - sponsored by Novartis. Phase 3
study of efficacy of amlodipine and aliskerin
combination therapy - to my knowledge this
combo is not available in the UK.
 President of the British Hypertension Society
2009-2010
 Member of the European Society of Hypertension
Council. Co-author on recent ESH Guideline reappraisal.
 I have no shares or interests in companies
marketing, or who have marketed therapies that
might be covered in this guidance.
None required.
 None made.
None required.
John Crimmins
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
Mark Caulfield
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
Michaela Watts
1st GDG meeting
(7th June 2010) until submission
318
Hypertension (partial update)
Declarations of Interest
GDG meeting
of guidance (21 January 2011)
Declaration
Action
 None made.
None required.
 None made.
None required.
 I have received speaker honoraria or attended
advisory boards for the following companies:
Boehringer-Ingelheim, Pfizer, Servier, SanofiAventis, Solvay, Takeda I have not had any further
pecuniary interests since that time and I will not
have any further hypertension-related pecuniary
interests in the 18 months of the GDG
membership.
None required.
Naomi Stetson
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
Shelley Mason
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
Terry McCormack
1st GDG meeting
(7th June 2010) until submission
of guidance (21 January 2011)
3rd GDG meeting
th
4 GDG meeting
6th GDG meeting
Declared on 8th January, 2011
 I am an investigator for Siguity which is a phase 3
study of Ivabradine, a Servier drug, which has no
relationship with hypertension. I received
sponsorship from merck, Sharpe and Dohne to
attend the European Society of Cardiology
meeting.
 I have received speaker fees from Merck to give
talks on lipids. I received an educational grant
from Boehringer Ingelheim to attend the annual
scientific meeting of the Primary Care
Cardiovascular Society.
 In relation to hypertension I have received
speaker’s fees for two talks given at Pulse
Magazine hypertension seminars sponsored by
Daiichi Sankyo.
 Not related to hypertension - I have received
advisory board honoraria, research grants,
speaker’s fees and educational grants from Astra
Zeneca, Boehringer Ingelheim, Merck Sharpe and
Dohme, Pfizer, Roche, and Servier.
 I was an investigator and steering committee
member for the HYVET study
 I am a member of the Primary Care Cardiovascular
Society Council.
 I am a member of the Executive Committee of the
British Hypertension Society.
 I am an Editor of the British Journal of
Cardiology.
Richard McManus
2nd GDG meeting (5 July 2010)
until submission of guidance (21
January 2011)
 None received
319
None required.
Hypertension (partial update)
Review questions
GDG meeting
Declared on 11th January, 2011
Declaration
Action
 I hold an NIHR Career Development Fellowship
which pays 75% of my salary. I run a research
group which holds several grants investigating
different methods of blood pressure monitoring
and the management of hypertension.
 I am a member of the NICE expert panel for the
Quality and Outcomes Framework for
hypertension and was previously a member of the
same panel prior to NICE taking over the QOF
contract.
 I am a member of the British Hypertension Society
and sit on its monitoring working group (Chair
from January 2011).
Appendix C: Review questions
Review questions
1.
In adults with suspected primary hypertension, what is the best method to measure blood pressure
(home vs ambulatory vs office) to establish the diagnosis and predict the development of CV events?
2.
In adults with treated primary hypertension, what is the best method to measure blood pressure
(home vs ambulatory vs office) for response to treatment and to predict the development of CV
events?
3.
In adults with primary hypertension, what protocol should be used when measuring ambulatory BP
for treatment and diagnosis?
4.
In adults with primary hypertension, what protocol should be used when measuring BP at home for
treatment and diagnosis?
5.
In adults with primary hypertension, at what BP should treatment be initiated?
6.
In adults with primary hypertension, what is the optimum BP that should be reached for once
treatment has been initiated/ targeted for treatment?
7.
If used, should ambulatory or home blood pressure readings be interpreted differently to office
measurements? i.e. are different thresholds for intervention/targets for treatment required, or should
adjustment be made to readings.
8.
In adults with primary hypertension, which is the most clinically and cost effective anti-hypertensive
monotherapy (ACEi vs ARB) for first-line treatment, and does this vary with age and ethnicity?
9.
In adults with primary hypertension, which is the most clinically and cost effective thiazide diuretic
(bendrofluazide / bendroflumethiazide, chlorthalidone, indapamide, hydrochlorothiazide) for firstline treatment, and does this vary with age and ethnicity?
10. In adults with primary hypertension, which is the most clinically and cost effective combination of
anti-hypertensives (A+C or A+D) for second line treatment, and does this vary with age and
ethnicity?
11. In adults with resistant hypertension, which is the most clinically and cost effective fourth-line
pharmacological treatment, and does this vary with age and ethnicity?
12. In adults with primary hypertension, what is the most clinically and cost effective first-line antihypertensive treatment (drug classes) in elderly people (aged ≥80 years)?
13. In adults with primary hypertension, what is the most clinically and cost effective first-line antihypertensive treatment (drug classes) in black people of African or Caribbean descent)?
320
Hypertension (partial update)
Literature search strategies
Appendix D: Literature search strategies
Search strategies used for the Hypertension guideline (partial update) are outlined below and
were run as per the NICE Guidelines Manual 2009
http://www.nice.org.uk/media/5F2/44/The_guidelines_manual_2009_-_All_chapters.pdf .
Searches for the clinical reviews were run in Medline (OVID), Embase (OVID), the
Cochrane Library (Wiley) and Cinahl (EBSCO). Usually, searches were constructed in the
following way:
A PICO format was used for intervention searches where population (P) terms were
combined with Intervention (I) and sometimes Comparison (C) terms. An intervention can be
a drug, a procedure or a diagnostic test. Outcomes (O) are rarely used in search strategies for
interventions. Search Filters were also added to the search where appropriate.
All search results retrieved on Cinahl were restricted to exclude Medline records. The
majority of searches were run from the final search date in either the original hypertension
guideline (CG18), i.e. 2003 or the rapid update (CG34), i.e. 2005. Searches run for all
available dates followed the dates in the table below. Full date parameters for each search are
given by question in section 1.3 of this appendix.
Database
Date available from
Medline
1950
Embase
1980
Cinahl
1982
The Cochrane Library
(to 2010 Issue 4)
1996 for Cochrane Reviews
1995 for DARE
1898 for CENTRAL
1904 for Methods Studies
1995 for HTA and NHSEED
Searches for the health economic reviews were run in Medline (Ovid), Embase (Ovid), the
NHS Economic Evaluations Database (NHS EED), the Health Technology Appraisals
database (HTA) and the Health Economic Evaluation Database (HEED). HTA and NHSEED
searches were carried out via the interface provided by the Centre for Reviews and
Dissemination (CRD). Searches of HTA, NHS EED and HEED were constructed using basic
population terms only. The economics databases were searched from 2003 onwards to find
anything published since the original guideline. There were two questions not covered in
either the original guideline or the previous rapid update, for which additional searches with
no date restrictions were carried out. Additionally, the search was run on MEDLINE and
Embase, with a specific economic filter, from 2009, For Medline and Embase an economic
filter (instead of a study type filter) was added to the same population terms used for the
clinical searches.
The final search date for all searches was 29th November 2010. Any studies added to the
databases after this date were not included unless specifically stated in the text.
The search strategies are presented below in the following order:
Section 1.1
Population terms by database. The same searches were used for all questions and for both
clinical and health economic searches.
Section 1.2
Study filter terms by database. These include filters for epidemiological study designs, health
economic studies and exclusions.
1.2.1
Systematic reviews
1.2.2
Randomised controlled trials (RCTs)
1.2.3
Combined RCT and Observational studies
321
Hypertension (partial update)
Literature search strategies
1.2.4
Health economics
1.2.5
Excluded study designs and publication types
1.2.6
Excluded population
Section 1.3
Searches run for specific questions with the intervention or exposure terms by database.
1.3.1
Measuring blood pressure
 AMB1
 AMB2
 AMB3
 AMB4
 AMB12
1.3.2
Thresholds for intervention
 THRESHOLD5
1.3.3
Targets for treatment
 OPTIMUM6
1.3.4
Pharmacological interventions
 ACEi vs ARB7
 DIURETICS8
 COMBI9
 RESISTANT10
 OVER80s11
 ETHNICITY13
D.1
Population search strategies
Medline
1
exp Hypertension/
2
hypertens*.ti,ab.
3
(essential adj3 hypertension).ti,ab.
4
(isolat* adj3 hypertension).ti,ab.
5
(elevat* adj3 blood adj pressur*).ti,ab.
6
(high adj3 blood adj pressur*).ti,ab.
7
(increase* adj3 blood pressur*).ti,ab.
8
((systolic or diastolic or arterial) adj3 pressur*).ti,ab.
9
or/1-8
10
exp pregnancy/
11
exp Hypertension, Pregnancy-Induced/
12
(pre eclampsia or pre-eclampsia or preeclampsia).ti,ab.
13
exp Hypertension, Malignant/
14
exp Hypertension, Portal/
15
exp Hypertension, Pulmonary/
16
exp Hypertension, Renal/
17
exp Intracranial Hypertension/
18
exp Ocular Hypertension/
19
exp diabetes mellitus/
20
or/10-19
322
Hypertension (partial update)
Literature search strategies
9 not 20
Embase
exp Hypertension/
1
hypertens*.ti,ab.
2
(essential adj3 hypertension).ti,ab.
3
(isolat* adj3 hypertension).ti,ab.
4
(elevat* adj3 blood adj pressur*).ti,ab.
5
(high adj3 blood adj pressur*).ti,ab.
6
(increase* adj3 blood pressur*).ti,ab.
7
((systolic or diastolic or arterial) adj3 pressur*).ti,ab.
8
or/1-8
9
exp pregnancy/
10
exp Hypertension, Pregnancy-Induced/
11
(pre eclampsia or pre-eclampsia or preeclampsia).ti,ab.
12
exp Hypertension, Malignant/
13
exp Hypertension, Portal/
14
exp Hypertension, Pulmonary/
15
exp Hypertension, Renal/
16
exp Intracranial Hypertension/
17
exp Ocular Hypertension/
18
exp diabetes mellitus/
19
or/10-19
20
9 not 20
323
Hypertension (partial update)
Literature search strategies
Cinahl
S1
S6 not S17
S17
S7 or S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16
S16
pregnancy n1 hypertension or malignant n1 hypertension or portal n1 hypertension or
pulmonary n1 hypertension or renal n1 hypertension or intracranial n1 hypertension or ocular
n1 hypertension
S15
pre eclampsia or pre-eclampsia or preeclampsia
S14
MH "ocular hypertension"
S13
MH "Intracranial hypertension"
S12
MH "Hypertension, Renal"
S11
MH "Hypertension, Portal"
S10
MH "Hypertension, Malignant"
S9
MH "Pregnancy-Induced Hypertension"
S8
(MH "Diabetes Mellitus+")
S7
MH "Pregnancy"
S6
S1 or S2 or S3 or S4 or S5
S5
essential n3 hypertension or isolat* n3 hypertension
S4
systolic n3 pressur* or diastolic n3 pressur* or arterial n3 pressur*
S3
high n3 blood n1 pressur* or increas* n3 blood n1 pressur* or elevat* n3 blood n1 pressur*
S2
hypertens*
S1
MH "hypertension"
Cochrane
#1
MeSH descriptor Hypertension explode all trees
#2
hypertens*:ti,ab,kw
#3
(essential or isolat*) near/3 hypertension:ti,kw,ab
#4
(elevat* or high or increase*) near/3 (blood near/1 pressur*):ti,ab,kw
#5
(systolic or diastolic or arterial) near/3 pressur*:ti,ab,kw
#6
(#1 OR #2 OR #3 OR #4 OR #5)
#7
MeSH descriptor Pregnancy explode all trees
#8
MeSH descriptor Hypertension, Pregnancy-Induced explode all trees
#9
MeSH descriptor Hypertension, Malignant explode all trees
#10
MeSH descriptor Hypertension, Portal explode all trees
#11
MeSH descriptor Hypertension, Pulmonary explode all trees
#12
MeSH descriptor Hypertension, Renal explode all trees
#13
MeSH descriptor Intracranial Hypertension explode all trees
#14
MeSH descriptor Ocular Hypertension explode all trees
#15
Pre eclampsia or pre-eclampsia or preeclampsia:ti.ab.kw
#16
(Pregnancy or malignant or portal or pulmonary or renal or intracranial or ocular) near/1
hypertension:ti,ab,kw
#17
MeSH descriptor Diabetes Mellitus explode all trees
#18
(#7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17)
#19
(#6 AND NOT #18)
324
Hypertension (partial update)
Literature search strategies
HTA / NHS EED – Centre for Reviews and Dissemination
#1
hypertens*
HEED (Health Economic Evaluations Database
#1
hypertens* in Article Title, Abstract or Keywords
D.2
Study design search terms
D.2.1
Systematic review search terms
Medline
1
review.pt. or review.ti. or "review"/
2
(systematic$ or evidence$ or methodol$ or quantitativ$ or analys$ or assessment$).ti,sh,ab.
3
1 and 2
4
meta-analysis.pt.
5
meta-analysis/
6
meta-analysis as topic/
7
"systematic review"/
8
(meta-analy$ or metanaly$ or metaanaly$ or meta analy$).ti,ab.
9
((systematic$ or evidence$ or methodol$ or quantitativ$) adj5 (review$ or survey$ or
overview$)).ti,ab,sh.
10
((pool$ or combined or combining) adj2 (data or trials or studies or results)).ti,ab.
11
or/3-10
Embase
1
review.pt. or review.ti. or "review"/
2
(systematic$ or evidence$ or methodol$ or quantitativ$ or analys$ or assessment$).ti,sh,ab.
3
1 and 2
4
meta-analysis.pt.
5
meta-analysis/
6
meta-analysis as topic/
7
"systematic review"/
8
(meta-analy$ or metanaly$ or metaanaly$ or meta analy$).ti,ab.
9
((systematic$ or evidence$ or methodol$ or quantitativ$) adj5 (review$ or survey$ or
overview$)).ti,ab,sh.
10
((pool$ or combined or combining) adj2 (data or trials or studies or results)).ti,ab.
11
or/45-52
325
Hypertension (partial update)
Literature search strategies
D.2.2
Randomised controlled trial (RCTs) search terms
Medline
1
randomized controlled trial$.pt,sh.
2
clinical trial$.pt,sh.
3
random allocation/
4
double blind method/
5
single blind method/
6
((clin$ or control$) adj5 trial$).ti,ab.
7
((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).ti,ab.
8
placebos/
9
placebo$.ti,ab.
10
random$.ti,ab.
11
(volunteer$ or "control group" or controls or prospective$).ti,ab.
12
research design/
13
or/1-12
14
animals/ not humans/
15
13 not 14
Embase
1
exp randomized controlled trial/
2
(random$ or placebo$).ti,ab.
3
((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).ti,ab.
4
double blind method/
5
exp comparative study/
6
exp evaluation/
7
exp follow up/
8
exp prospective study/
9
placebos/
10
or/1-9
11
exp human/
12
10 and 11
D.2.3
Combined RCT and Observational studies search terms
Medline
1
randomized controlled trial.pt.
2
controlled clinical trial.pt.
3
double-blind method/ or random allocation/ or single-blind method/
4
exp Clinical Trial/
5
exp Clinical Trials as Topic/
6
clinical trial.pt.
7
random.ti,ab.
8
(clin$ adj25 trial$).ti,ab.
9
((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).ti,ab.
326
Hypertension (partial update)
Literature search strategies
10
Placebos/ or placebo$.ti,ab.
11
Research Design/ or Comparative Study/
12
exp Evaluation Studies/ or follow-up studies/ or prospective studies/
13
(volunteer$ or "control group" or controls or prospectiv$).ti,ab.
14
exp epidemiological studies/
15
cohort stud$.ti,ab.
16
case control stud$.ti,ab.
17
((crossover or cross-over or cross over) adj2 (design$ or stud$ or procedure$ or trial$)).ti,ab.
18
or/1-17
Embase
1
controlled study/ or randomized controlled trial/
2
Clinical Trial/
3
clinical study/ or major clinical study/ or clinical trial/ or phase 1 clinical trial/ or phase 2
clinical trial/
4
Placebo/
5
"Double Blind Procedure"/
6
Randomization/
7
((clinical$ or control$ or compar$) adj3 (trial$ or study or studies)).mp.
8
or/1-7
9
compar$.tw.
10
control$.tw.
11
9 and 10
12
placebo.tw.
13
randomi$.tw.
14
(blind$ or mask$).tw.
15
crossover procedure/
16
(cross adj2 over adj2 (study or design)).ti,ab.
17
exp Cohort Analysis/
18
exp Longitudinal Study/
19
exp Prospective Study/
20
Observational Study/
21
exp follow up/
22
cohort studies.ti,ab.
23
exp Case Control Study/
24
case control stud$.ti,ab.
25
or/8,11,12-24
D.2.4
Health economics search terms
Medline
1
exp "Costs and Cost Analysis"/
2
economics/
3
exp Economics, Hospital/
4
exp Economics, Medical/
5
exp Economics, Nursing/
327
Hypertension (partial update)
Literature search strategies
6
exp Economics, Pharmaceutical/
7
exp "Fees and Charges"/
8
exp Budgets/
9
ec.fs.
10
(economic$ or pharmacoeconomic$ or price$ or pricing$ or cost$ or budget$).ti,ab.
11
(value adj2 (money or monetary)).ti,ab.
12
(expenditure not energy).ti,ab.
13
or/1-12
14
((metabolic or energy or oxygen) adj1 cost$).ti,ab.
15
13 not 14
Embase
1
health economics/
2
exp economic evaluation/
3
exp health care cost/
4
exp pharmacoeconomics/
5
exp fee/
6
budget/
7
(economic$ or pharmacoeconomic$ or cost$ or price$ or pricing$ or budget$).ti,ab.
8
(value adj2 (money or monetary$)).ti,ab.
9
(expenditure not energy).ti,ab.
10
or/1-9
11
((metabolic or energy or oxygen) adj1 cost$).ti,ab.
12
10 not 11
D.2.5
Excluded study designs search terms
The following study designs and publication types were removed from the retrieved results using the NOT
operator.
Medline & Embase
1
letter$/
2
editorial.pt.
3
historical article.pt.
4
anecdote.pt.
5
commentary.pt.
6
note.pt.
7
case report/
8
case report$.pt.
9
case study/
10
case study.pt.
11
exp animal/ not human/
12
nonhuman/
13
exp animal studies/
14
animals, laboratory/
328
Hypertension (partial update)
Literature search strategies
15
exp experimental animal/
16
exp animal experiment/
17
exp animal model/
18
exp rodentia/
19
exp rodents/
20
exp rodent/
21
or/1-20
D.2.6
Excluded population search terms
The following populations were removed from the retrieved results using the NOT operator.
Medline
1
exp child/
2
child$.tw.
3
exp infant/
4
infan$.tw.
5
(baby or babies).tw.
6
"adolescent"/
7
(pediatric$1 or paediatric$1).tw.
8
or/1-7
9
exp adult/
10
8 not 9
Embase
1
exp child/
2
child*.tw.
3
childhood/
4
infancy/
5
infan*.tw.
6
(baby or babies).tw.
7
exp adolescent/
8
(pediatric$1 or paediatric$1).tw.
9
or/1-8
10
adult/
11
aged/
12
or/10-11
13
9 not 12
329
Hypertension (partial update)
Literature search strategies
D.3
Searches by specific question
D.3.1
Measuring blood pressure
AMB1
In adults with suspected primary hypertension, what is the best method to measure blood
pressure (home vs ambulatory vs office) to establish the diagnosis and predict the
development of CV events?
AMB2
In adults with treated primary hypertension, what is the best method to measure blood
pressure (home vs ambulatory vs office) for response to treatment and to predict the
development of CV events?
Search constructed by combining the columns in the following table using the AND Boolean
operator
Intervention /
exposure
Population
Adults with
primary
hypertension
Home or
ambulatory or
office blood
pressure
measurement
Comparison
Study filter used
Date parameters
None
SRs / Combined
RCT and
Observational filter
(Medline and
Embase only)
2003-29/11/10
Medline
1
ambulatory blood pressure monitoring.ti,ab.
2
ambulat*.ti,ab.
3
((home or self or office) adj2 "'blood monitor*").ti,ab.
4
blood pressure monitor*.ti,ab.
5
24-hour.ti,ab.
6
24 hour monitor*.ti,ab.
7
or/1-6
8
exp "Sensitivity and Specificity"/
9
(sensitivity or specificity).ti.
10
(predictive adj3 value*).tw.
11
exp diagnostic errors/
12
((False adj positive*) or (false adj negative*)).ti.
13
(observer adj variations*).ti.
14
(roc adj curve*).ti.
15
(likelihood adj ratio*).ti.
16
Likelihood function/
17
or/8-16
18
exp *prognosis/
19
prognos*.ti.
20
prevalence/
21
incidence/
22
(prevalence or incidence).ti.
23
or/18-22
330
Hypertension (partial update)
Literature search strategies
24
7 and (17 or 23)
Embase
1
ambulatory blood pressure monitoring.ti,ab.
2
ambulat*.ti,ab.
3
((home or self or office) adj2 "'blood monitor*").ti,ab.
4
blood pressure monitor*.ti,ab.
5
24-hour.ti,ab.
6
24 hour monitor*.ti,ab.
7
or/1-6
8
exp "Sensitivity and Specificity"/
9
(sensitivity or specificity).ti.
10
(predictive adj3 value*).tw.
11
exp diagnostic errors/
12
((False adj positive*) or (false adj negative*)).ti.
13
(observer adj variations*).ti.
14
(roc adj curve*).ti.
15
(likelihood adj ratio*).ti.
16
Likelihood function/
17
or/8-16
18
exp *prognosis/
19
prognos*.ti.
20
prevalence/
21
incidence/
22
(prevalence or incidence).ti.
23
or/18-22
24
7 and (17 or 23)
Cinahl
S28
S10 and S27
S27
S20 or S26
S26
S21 or S22 or S23 or S24 or S25
S25
prevalence or incidence
S24
(MH "Incidence")
S23
(MH "Prevalence")
S22
prognos*
S21
(MH "Prognosis")
S20
S11 or S12 or S13 or S14 or S15 or S16 or S17 or S18 or S19
S19
"likelihood function*"
S18
likelihood n3 ratio*
S17
roc n1 curve*
S16
observer n1 variation*
S15
false n1 positive* or false n1 negative*
S14
(MH "Diagnostic Errors")
S13
predictive n3 value*
331
Hypertension (partial update)
Literature search strategies
S12
sensitivity or specificity
S11
(MH "Sensitivity and Specificity")
S10
S1 or S2 or S3 or S4 or S5 or S6 or S7 or S8 or S9
S9
24 hour monitor*
S8
24-hour
S7
self and blood and monitor*
S6
office and blood and monitor*
S5
home and blood and monitor*
S4
blood pressure monitor*
S3
ambulat*
S2
ambulatory blood pressure monitoring
S1
MH "Blood Pressure Monitoring, Ambulatory"
Cochrane
#1
MeSH descriptor Blood Pressure Monitoring, Ambulatory explode all trees
#2
ambulatory blood pressure monitoring:ti,kw,ab
#3
ambulat*:ti,kw,ab
#4
blood pressure monitor*:ti,kw,ab
#5
(home or self or office) near/2 (blood near/1 monitor*):ti,kw,ab
#6
24-hour:ti,kw,ab
#7
24 hour monitor*:ti,kw,ab
#8
(#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7)
#9
MeSH descriptor Sensitivity and Specificity explode all trees
#10
sensitivity:ti or specificity:ti,kw,ab
#11
predictive near/3 value*:ti,kw,ab
#12
MeSH descriptor Diagnostic Errors explode all trees
#13
(false near/1 positive*) or (false near/1 negative*):ti,kw,ab
#14
observer near/1 variation*:ti,kw,ab
#15
roc near/1 curve*:ti,ab,kw
#16
(likelihood near/3 ratio*):ti,ab,kw
#17
MeSH descriptor Likelihood Functions explode all trees
#18
(#9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 or #17)
#19
MeSH descriptor Prognosis explode all trees
#20
prognos*:ti,ab,kw
#21
MeSH descriptor Prevalence explode all trees
#22
MeSH descriptor Incidence explode all trees
#23
(prevalence or incidence):ti,ab,kw
#24
(#19 OR #20 OR #21 OR #22 OR #23)
#25
(#18 OR #24)
#26
(#8 AND #25)
AMB3
In adults with primary hypertension, if used, should ambulatory or home blood pressure
readings be interpreted differently to office measurements? i.e. are different thresholds for
intervention/targets for treatment required, or should adjustment be made to readings.
332
Hypertension (partial update)
Literature search strategies
AMB4
In adults with primary hypertension, what protocol should be used when measuring BP at
home for treatment and diagnosis?
Search constructed by combining the columns in the following table using the AND Boolean
operator
Intervention /
exposure
Home or
ambulatory blood
pressure
monitoring
Population
Adults with
primary
hypertension
Comparison
Study filter used
Date parameters
None
None
2003-29/11/10
Medline
1
*Blood Pressure Monitoring, Ambulatory/
2
((blood pressure* or BP) adj3 ambulatory).ti,ab.
3
((blood pressure* or BP) adj3 home).ti,ab.
4
((blood pressure* or BP) adj3 self).ti,ab.
5
HBPM.ti,ab.
6
or/1-5
Embase
((blood pressure* or BP) adj3 home).ti,ab.
((blood pressure* or BP) adj3 ambulatory).ti,ab.
((blood pressure* or BP) adj3 self).ti,ab.
HBPM.ti,ab.
*blood pressure monitoring/
or/1-5
Cinahl
S6
S1 or S2 or S3 or S4 or S5
S5
HBPM
S4
((blood pressure* or BP) and self)
S3
((blood pressure* or BP) and ambulatory)
S2
((blood pressure* or BP) and home)
S1
MM "Blood Pressure Monitoring, Ambulatory"
Cochrane
#1
MeSH descriptor Blood Pressure Monitoring, Ambulatory, this term only
#2
Ambulatory blood pressure monitoring:ti,kw,ab
#3
((blood pressure* or BP) near/3 home):ti,kw,ab
#4
((blood pressure* or BP) near/3 self):ti,kw,ab
#5
((blood pressure* or BP) near/3 ambulatory):ti,kw,ab
#6
HBPM:ti,kw,ab
#7
(#1 OR #2 OR #3 OR #4 OR #5 OR #6)
333
Hypertension (partial update)
Literature search strategies
AMB12
In adults with primary hypertension, what protocol should be used when measuring
ambulatory blood pressure for treatment and diagnosis?
Search constructed by combining the columns in the following table using the AND Boolean
operator
Intervention /
exposure
Ambulatory blood
pressure
monitoring
Population
Adults with
primary
hypertension
Comparison
Study filter used
Date parameters
None
None
All dates-29/11/10
Medline
1
*Blood Pressure Monitoring, Ambulatory/
2
((blood pressure* or BP) adj ambulatory).ti,ab.
3
ABPM.ti,ab.
4
or/1-3
Embase
1
((blood pressure* or BP) adj ambulatory).ti,ab.
2
ABPM.ti,ab.
3
*blood pressure monitoring/
4
or/1-3
Cinahl
S4
S1 or S2 or S3
S3
ABPM
S2
((blood pressure* or BP) and ambulatory)
S1
MM "Blood Pressure Monitoring, Ambulatory"
Cochrane
#1
MeSH descriptor Blood Pressure Monitoring, Ambulatory, this term only
#2
(ambulatory near (blood pressure or BP)):ti,kw,ab
#3
ABPM:ti,kw,ab
#4
(#1 OR #2 OR #3)
D.3.2
Thresholds for intervention
THRESHOLD5
In adults with primary hypertension, at what BP should treatment be initiated?
Search constructed by combining the columns in the following table using the AND Boolean
operator
Population
Adults with
primary
hypertension
Intervention /
exposure
Comparison
Study filter used
Date parameters
Any treatment
None
None
2003-29/11/10
334
Hypertension (partial update)
Literature search strategies
Medline
1
exp antihypertensive agents/
2
(anti-hypertens* or antihypertens* or anti hypertens*).ti,ab.
3
or/1-2
4
risk factors/
5
risk assessment/
6
or/4-5
7
3 and 6
Embase
1
exp antihypertensive agent/
2
exp antihypertensive therapy/
3
(anti-hypertens* or antihypertens* or anti hypertens*).ti,ab.
4
or/1-3
5
risk factor/
6
risk assessment/
7
or/5-6
8
exp blood pressure/
9
blood pressure monitoring/
10
or/8-9
11
4 and 7 and 10
Cinahl
S7
S3 and S6
S6
S4 or S5
S5
(MH "Risk Assessment")
S4
(MH "Risk Factors+")
S3
S1 or S2
S2
(MH "Antihypertensive Agents+")
S1
anti-hypertens* or antihypertens* or anti hypertens*
Cochrane
#1
(anti-hypertens* or antihypertens* or anti hypertens*):ti,ab,kw
#2
risk*:ti,ab,kw
#3
(#1 AND #2)
D.3.3
Targets for treatment
OPTIMUM6
In adults with primary hypertension, what is the optimum BP that should be reached for once
treatment has been initiated/ targeted for treatment?
Population
Adults with
primary
Intervention /
exposure
Comparison
Study filter used
Date parameters
Any treatment
None
None
2003-29/11/10
335
Hypertension (partial update)
Literature search strategies
Intervention /
exposure
Population
hypertension
Comparison
Study filter used
Date parameters
Medline
1
((target* or level* or optimum or optimal) adj3 blood pressure).ti,ab.
2
(intensive adj3 (blood pressure or anti?hypertens* or control or treatment)).ti,ab.
3
((tight* or strict*) adj3 (blood pressure or control*)).ti,ab.
4
(normotensive* or normotension).ti,ab.
5
or/1-4
6
*blood pressure/
7
5 and 6
Embase
1
((target* or level* or optimum or optimal) adj3 blood pressure).ti,ab.
2
(intensive adj3 (blood pressure or anti?hypertens* or control or treatment)).ti,ab.
3
((tight* or strict*) adj3 (blood pressure or control*)).ti,ab.
4
(normotensive* or normotension).ti,ab.
5
or/1-4
6
exp blood pressure/
7
5 and 6
Cinahl
S7
S5 and S6
S6
(anti-hypertens* or antihypertens* or anti hypertens*)
S5
S1 or S2 or S3 or S4
S4
(normotensive* or normotension)
S3
((tight* or strict*) and (blood pressure or control*))
S2
(intensive and (blood pressure or antihypertens* or anti-hypertens* or anti hypertens* or control
or treatment))
S1
((target* or level* or optimum or optimal) and blood pressure)
Cochrane
#1
((target* or level* or optimum or optimal) near/3 blood pressure):ti,ab,kw
#2
(intensive near/3 (blood pressure or anti hypertens* or anti-hypertens* or antihypertens* or
control or treatment)):ti,ab,kw
#3
((tight* or strict*) near/3 (blood pressure or control*)):ti,ab,kw
#4
(normotensive* or normotension):ti,ab,kw
#5
(#1 OR #2 OR #3 OR #4)
#6
(anti-hypertens* or antihypertens* or anti hypertens*):ti,ab,kw
#7
(#5 AND #6)
D.3.4
Pharmacological interventions
ACEivsARB7: In adults with primary hypertension, which is the most clinically and cost
effective anti-hypertensive monotherapy (ACEi vs ARB) for first-line treatment and does this
vary by age and ethnicity?
336
Hypertension (partial update)
Literature search strategies
Intervention /
exposure
ACE inhibitor
Population
Adults with
primary
hypertension
Comparison
Study filter used
Date parameters
ARB
SRs / RCTs
Medline and
Embase only
2005-29/11/10
Medline
1
*angiotensin converting enzyme inhibitors/
2
ace inhibitor*.ti,ab.
3
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
4
or/1-3
5
*Angiotensin II Type 1 Receptor blockers/
6
*Angiotensin 2 receptor antagonist/
7
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
8
or/5-7
9
4 and 8
Embase
1
*Angiotensin II Type 1 Receptor blockers/
2
*Angiotensin 2 receptor antagonist/
3
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
4
or/1-3
5
*angiotensin converting enzyme inhibitors/
6
ace inhibitor*.ti,ab.
7
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
8
or/5-7
9
4 and 8
Cinahl
S7
S3 and S6
S6
S4 or S5
S5
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril
S4
MH "Angiotensin-Converting Enzyme Inhibitors"
S3
S1 or S2
S2
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or olmesartan
S1
MH "Angiotensin II Type I Receptor Blockers"
Cochrane
#1
MeSH descriptor Angiotensin II Type 1 Receptor Blockers explode all trees
#2
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan:ti,ab,kw
#3
(#1 OR #2)
#4
MeSH descriptor Angiotensin-Converting Enzyme Inhibitors explode all trees
337
Hypertension (partial update)
Literature search strategies
#5
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril:ti,ab,kw
#6
(#4 OR #5)
#7
(#3 AND #6)
DIURETICS8: In adults with primary hypertension, which is the most clinically and cost
effective thiazide diuretic (bendrofluazide/bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide) for first line treatment and does this vary by age and ethnicity?
Intervention /
exposure
Thiazide diuretic
Population
Adults with
primary
hypertension
Comparison
Study filter used
Date parameters
None
SRs / RCTs
Medline and
Embase only
All dates-29/11/10
Medline
1
bendroflumethiazide/
2
chlorthalidone/
3
indapamide/
4
metolazone/
5
xipamide/
6
hydrochlorothiazide/
7
exp Sodium Chloride Symporter Inhibitors/
8
diuretics/
9
7 and 8
10
thiazide diuretic*.ti,ab.
11
or/1-6,9-10
Embase
1
*bendroflumethiazide/
2
*chlortalidone/
3
*indapamide/
4
*metolazone/
5
*xipamide/
6
*hydrochlorothiazide/
7
*thiazide diuretic agent/
8
thiazide diuretic*.ti,ab.
9
or/1-8
Cinahl
S4
S1 or S2 or S3
S3
bendroflumethiazide or chlorthalidone or indapamide or metolazone or xipamide or
hydrochlorothiazide
S2
thiazide diuretic*
S1
(MH "Diuretics, Thiazide+")
Cochrane
#1
(bendroflumethiazide or chlorthalidone or indapamide or metolazone or xipamide or
hydrochlorothiazide):ti,ab,kw
338
Hypertension (partial update)
Literature search strategies
#2
thiazide diuretic*:ti,ab,kw
#3
MeSH descriptor Sodium Chloride Symporter Inhibitors explode all trees
#4
(#1 OR #2)
#5
MeSH descriptor Diuretics, this term only
#6
(#3 AND #5)
#7
(#4 OR #6)
COMBI9: In adults with primary hypertension, which is the most clinically and cost effective
combination of anti-hypertensives (A+C or A+D) for second line treatment and does this vary
by age and ethnicity?
Intervention /
exposure
Population
Adults with
primary
hypertension
ACE inhibitor +
Calcium Channel
Blocker (CCB) or
ARB + CCB in
second line
treatment
Comparison
Study filter used
Date parameters
None
SRs / RCTs
Medline and
Embase only
2005-29/11/10
Medline
1
*angiotensin converting enzyme inhibitors/
2
ace inhibitor*.ti,ab.
3
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
4
or/1-3
5
*Angiotensin II Type 1 Receptor blockers/
6
*Angiotensin 2 receptor antagonist/
7
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
8
or/5-7
9
4 or 8
10
*calcium channel blockers/
11
(amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or
nifedipine or nimodipine or diltiazem or verapamil).ti,ab.
12
10 or 11
13
9 and 12
Embase
1
*angiotensin converting enzyme inhibitors/
2
ace inhibitor*.ti,ab.
3
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
4
or/1-3
5
*Angiotensin II Type 1 Receptor blockers/
6
*Angiotensin 2 receptor antagonist/
7
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
8
or/5-7
9
4 or 8
339
Hypertension (partial update)
Literature search strategies
10
*calcium channel blockers/
11
(amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or
nifedipine or nimodipine or diltiazem or verapamil).ti,ab.
12
10 or 11
13
9 and 12
Cinahl
S11
S17 and S10
S10
S8 or S9
S9
amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or nifedipine
or nimodipine or diltiazem or verapamil
S8
MH "calcium channel blockers"
S7
S3 or S6
S6
S4 or S5
S5
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril
S4
MH "Angiotensin-Converting Enzyme Inhibitors"
S3
S1 or S2
S2
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or olmesartan
S1
MH "Angiotensin II Type I Receptor Blockers"
Cochrane
#1
MeSH descriptor Angiotensin II Type 1 Receptor Blockers explode all trees
#2
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan:ti,ab,kw
#3
(#1 OR #2)
#4
MeSH descriptor Angiotensin-Converting Enzyme Inhibitors explode all trees
#5
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril:ti,ab,kw
#6
(#4 OR #5)
#7
(#3 OR #6)
#8
MeSH descriptor Calcium Channel Blockers explode all trees
#9
(amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or nifedipine
or nimodipine or diltiazem or verapamil):ti,ab,kw
#10
(#8 OR #9)
#11
(#7 AND #10)
RESISTANT10: In adults with resistant hypertension, which is the most clinically and cost
effective fourth-line pharmacological treatment and does this vary by age and ethnicity?
Population
Adults with
primary
hypertension
Intervention /
exposure
Any fourth line
pharmacological
treatment
Comparison
Study filter used
Date parameters
None
SRs / RCTs
Medline and
Embase only
2005-29/11/10
Medline
1
((difficult* or hard or poor*) adj2 control*).ti,ab.
2
(uncontrolled adj2 hypertens*).ti,ab.
3
(fourth line or fourth-line).ti,ab.
340
Hypertension (partial update)
Literature search strategies
4
drug resistance/
5
((resistant or resistance) adj3 hypertens*).ti,ab.
6
or/1-5
Embase
1
((difficult* or hard or poor*) adj2 control*).ti,ab.
2
(uncontrolled adj2 hypertens*).ti,ab.
3
(fourth line or fourth-line).ti,ab.
4
((resistant or resistance) adj3 hypertens*).ti,ab.
5
drug resistance/
6
or/1-5
Cinahl
S6
S1 or S2 or S3 or S4 or S5
S5
(MH "Drug Resistance")
S4
((resistant or resistance) and hypertens*)
S3
(fourth line or fourth-line)
S2
(uncontrolled and hypertens*)
S1
((difficult* or hard or poor*) and control*)
Cochrane
#1
((difficult* or hard or poor*) near/2 control*):ti,ab,kw
#2
(uncontrolled near/2 hypertens*):ti,ab,kw
#3
(fourth line or fourth-line):ti,ab,kw
#4
((resistant or resistance) near/3 hypertens*):ti,ab,kw
#5
MeSH descriptor Drug Resistance, this term only
#6
(#1 OR #2 OR #3 OR #4 OR #5)
OVER80s11: Does the treatment of hypertension in people aged ≥ 80 years reduce the risk of
major adverse outcomes?
Population
Adults (≥80 years
old) with primary
hypertension
Intervention /
exposure
Any
pharmacological
treatment
Comparison
Study filter used
Date parameters
None
SRs / RCTs
Medline and
Embase only
2005-29/11/10
Medline
1
*diuretics, thiazide/
2
(chlortalidone or indapamide or metolazone or xipamide or amiloride or triamterene).ti,ab.
3
1 or 2
4
*adrenergic alpha-antagonists/
5
(doxazosin or indoramin or prazosin or terazosin).ti,ab.
6
4 or 5
7
*adrenergic beta-antagonists/
8
(Oxprenolol or pindolol or acebutolol or celiprolol or atenolol or nadolol or sotalol or carvedilol
or bisoprolol or metoprolol or nebivolol or timolol or propranolol or betaxalol or esmolol or
labetalol).ti,ab.
341
Hypertension (partial update)
Literature search strategies
9
7 or 8
10
*calcium channel blockers/
11
(amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or nifedipine
or nimodipine or diltiazem or verapamil).ti,ab.
12
10 or 11
13
*angiotensin converting enzyme inhibitors/
14
ace inhibitor*.ti,ab.
15
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
16
or/13-15
17
*Angiotensin II Type 1 Receptor blockers/
18
*Angiotensin 2 receptor antagonist/
19
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
20
or/17-19
21
3 or 6 or 9 or 12 or 16 or 20
22
(80 years or aged 80 or centenarian* or nonagenarian* or octogenarian* or elderly or
older).ti,ab.
23
exp aged/
24
22 or 23
25
21 and 24
Embase
1
*Angiotensin II Type 1 Receptor blockers/
2
*Angiotensin 2 receptor antagonist/
3
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
4
or/1-3
5
*angiotensin converting enzyme inhibitors/
6
ace inhibitor*.ti,ab.
7
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
8
or/5-7
9
*adrenergic alpha-antagonists/
10
(doxazosin or indoramin or prazosin or terazosin).ti,ab.
11
or/9-10
12
*adrenergic beta-antagonists/
13
(Oxprenolol or pindolol or acebutolol or celiprolol or atenolol or nadolol or sotalol or carvedilol
or bisoprolol or metoprolol or nebivolol or timolol or propranolol or betaxalol or esmolol or
labetalol).ti,ab.
14
12 or 13
15
*calcium channel blockers/
16
(amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or nifedipine
or nimodipine or diltiazem or verapamil).ti,ab.
17
15 or 16
18
*diuretics, thiazide/
19
(chlortalidone or indapamide or metolazone or xipamide or amiloride or triamterene).ti,ab.
20
or/18-19
342
Hypertension (partial update)
Literature search strategies
21
4 or 8 or 11 or 14 or 17 or 20
22
(80 years or aged 80 or centenarian* or nonagenarian* or octogenarian* or elderly or
older).ti,ab.
23
exp aged/
24
22 or 23
25
21 and 24
Cinahl
S19
S3 and S18
S18
S4 or S5 or S6 or S7 or S8 or S9 or S10 or S11 or S12 or S13 or S14 or S15 or S16 or S17
S17
(MH "Adrenergic Beta-Antagonist+")
S16
(MH "Adrenergic Alpha-Antagonists+")
S15
oxprenolol or pindolol or acebutolol or celiprolol or atenolol or nadolol or sotalol or carvedilol
or bisoprolol or metoprolol or nebivolol or timolol or propranolol or betaxalol or esmolol or
labetalol
S14
doxazosin or indoramin or prazosin or terazosin
S13
chlortalidone or indapamide or metolazone or xipamide or amiloride or triamterene
S12
bendroflumethiazide or chlorthalidone or indapamide or metolazone or xipamide
S11
(MH "Diuretics+")
S10
(MH "Diuretics, Thiazide+")
S9
amlodipine or felodipine or isradipine or lacidipine or lercanidipine or nicardipine or nifedipine
or nimodipine or diltiazem or verapamil
S8
MH "calcium channel blockers"
S7
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril
S6
MH "Angiotensin-Converting Enzyme Inhibitors"
S5
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or olmesartan
S4
MH "Angiotensin II Type I Receptor Blockers"
S3
S1 or S2
S2
80 years or aged 80 or centenarian* or nonagenarian* or octogenarian* or elderly or older
S1
(MH "Aged, 80 and Over")
Cochrane
#1
(80 years or aged 80 or centenarian* or nonagenarian* or octogenarian* or elderly or
older):ti,ab,kw
#2
MeSH descriptor Aged, 80 and over, this term only
#3
(#1 OR #2)
#4
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or olmesartan or
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril or amlodipine or felodipine or isradipine or
lacidipine or lercanidipine or nicardipine or nifedipine or nimodipine or diltiazem or
verapamil):ti,ab,kw
#5
(chlortalidone or indapamide or metolazone or xipamide or amiloride or triamterene or
doxazosin or indoramin or prazosin or terazosin or oxprenolol or pindolol or acebutolol or
celiprolol or atenolol or nadolol or sotalol or carvedilol or bisoprolol or metoprolol or nebivolol
or timolol or propranolol or betaxalol or esmolol or labetalol):ti,ab,kw
#6
(#4 OR #5)
#7
(#3 AND #6)
343
Hypertension (partial update)
Literature search strategies
ETHNICITY13: In adults with primary hypertension, what is the most clinically and cost
effective first-line anti-hypertensive treatment (drug classes) in black people of African or
Caribbean descent)?
Population
Black adults (≥18
years old) of
African or
Caribbean descent
with primary
hypertension
Intervention /
exposure
ACE inhibitors /
ARBs
Comparison
Study filter used
Date parameters
None
SRs / RCTs
Medline and
Embase only
2005-29/11/10
Medline
1
*angiotensin converting enzyme inhibitors/
2
ace inhibitor*.ti,ab.
3
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
4
or/1-3
5
*Angiotensin II Type 1 Receptor blockers/
6
*Angiotensin 2 receptor antagonist/
7
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
8
or/5-7
9
4 or 8
10
(black* or afro* or african*).ti,ab,hw.
11
exp population groups/
12
10 or 11
13
9 and 12
Embase
1
*Angiotensin II Type 1 Receptor blockers/
2
*Angiotensin 2 receptor antagonist/
3
(candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan).ti,ab.
4
or/1-3
5
*angiotensin converting enzyme inhibitors/
6
ace inhibitor*.ti,ab.
7
(captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril).ti,ab.
8
or/5-7
9
4 or 8
10
(black* or afro* or african*).ti,ab,hw.
11
exp "ethnic or racial aspects"/
12
exp "ethnic and racial groups"/
13
or/10-12
14
9 and 13
Cinahl
S11
S5 and S10
344
Hypertension (partial update)
Review protocols
S10
S6 or S7 or S8 or S9
S9
black* or afro* or african*
S8
(MH "Race Factors")
S7
(MH "Minority Groups")
S6
(MH "Ethnic Groups+")
S5
S1 or S2 or S3 or S4
S4
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or olmesartan
S3
MH "Angiotensin II Type I Receptor Blockers"
S2
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril
S1
MH "Angiotensin-Converting Enzyme Inhibitors"
Cochrane
#1
MeSH descriptor Angiotensin II Type 1 Receptor Blockers explode all trees
#2
candesartan or eprosartan or ibesartan or telmisartan or valsartan or losartan or
olmesartan:ti,ab,kw
#3
(#1 OR #2)
#4
MeSH descriptor Angiotensin-Converting Enzyme Inhibitors explode all trees
#5
captopril or cilazapril or enalapril or fosinopril or imidapril or lisinopril or moexipril or
perindopril or quinapril or ramipril or trandolapril:ti,ab,kw
#6
(#4 OR #5)
#7
(#3 OR #6)
#8
MeSH descriptor Population Groups explode all trees
#9
(black* or afro* or african*):ti,ab,kw
#10
(#8 OR #9)
#11
(#7 AND #10)
Appendix E: Review protocols
E.1
Review protocol for the diagnosis of hypertension and
monitoring treatment efficacy
 Predicting outcome using clinic, home and ambulatory measurements
 Sensitivity and specificity of clinic, home and ambulatory measurements
 Measuring response to treatment using clinic, home and ambulatory measurements
Review
questions
Objectives
AMB1: In adults with suspected primary hypertension, what is the best method to
measure blood pressure (home vs ambulatory vs office) to establish the diagnosis and
predict the development of CV events?
AMB2: In adults with treated primary hypertension, what is the best method to
measure blood pressure (home vs ambulatory vs office) for response to treatment and
to predict the development of CV events?
This aim of this review is to estimate the accuracy and cost-effectiveness of BP
measurements (home vs ambulatory vs. office) for establishing the diagnosis of and
prognosis in adults ≥18 years old with suspected primary hypertension; and for
establishing response to treatment in adults ≥18 years old with treated primary
345
Hypertension (partial update)
Review protocols
hypertension.
Criteria
Population(s):
AMB1: Adults ≥18 years old with suspected primary hypertension who may or may not
have pre-existing cardiovascular disease, chronic kidney disease and diabetes
AMB2: Adults ≥18 years old with treated primary hypertension who may or may not have
pre-existing cardiovascular disease, chronic kidney disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention(s):
Home; ambulatory (and home vs. ambulatory)
Studies looking at home telemonitoring BP will be excluded, as this not common in
clinical practice
Comparison(s):
Gold standard (office)
Outcome(s):
 Sensitivity
 Specificity
 Risk of developing clinical outcomes:
 Mortality
 Stroke
 MI Vascular procedures (including both coronary and arotid artery procedures)
 Angina requiring hospitalisation
Major adverse cardiac and cerebrovascular events (MAACE): fatal and non-fatal MI, fatal
and non-fatal stroke, hospitalised angina, hospitalised heart failure, revascularisation (AND
DIFFERENT COMPOSITES OF THIS OUTCOME)
Studies will be excluded if they:
 have no comparative BP measurement arm
 are validation studies
 are RCTs for treatment that have not monitored Tx throughout trial by ≥two different BP
measurement methods (office or home or ABPM) and use two methods only at end of study
 assess the prediction of future BP (rather than the specified clinical outcomes)
Search
Strategy
See Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG18, however the studies included within the SR/MA may
have been published pre- and post- 2003); RCTs; non-RCTs: mainly diagnostic, prognostic
and cohort studies
Taking into consideration the advice on prognostic reviews in the NICE guidelines manual,
meta-analysis or GRADE will not be undertaken for prognostic studies as well as for other
non-RCTs.
346
Hypertension (partial update)
Review protocols
E.2
Review protocols for the diagnosis of hypertension:
measurement protocols for diagnosing hypertension
 Ambulatory blood pressure measurement
 Home blood pressure measurement
Review
questions
Objectives
Criteria
AMBU PROTOCOL: In adults with primary hypertension, what protocol should be
used when measuring ambulatory BP for treatment and diagnosis?
HOME PROTOCOL: In adults with primary hypertension, what protocol should be
used when measuring BP at home for treatment and diagnosis?
This aim of this review is to establish the best protocols for measuring ambulatory BP and
home BP for diagnosis of hypertension and for monitoring treatment (in adults ≥18 years
old).
Population (s):
Adults ≥18 years old with suspected primary hypertension (diagnosis) or hypertension
(treatment) who may or may not have pre-existing cardiovascular disease, chronic kidney
disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s) and comparison(s):
Home vs Home; ABPM vs ABPM
Studies looking at home telemonitoring BP will be excluded, as this not common in
clinical practice
Studies will be selected from abstract lists if they mention more than 1 method or protocol
of measuring home BP or if they directly assess what the optimum HBP protocol should
be; otherwise they will be excluded at the abstracting stage.
Outcome(s):
 Sensitivity
 Specificity
 Reliability / reproducibility







Risk of developing clinical outcomes:
Mortality
Stroke
MI
Vascular procedures (including both coronary and arotid artery procedures)
Angina requiring hospitalisation
Major adverse cardiac and cerebrovascular events (MAACE): fatal and non-fatal MI, fatal
and non-fatal stroke, hospitalised angina, hospitalised heart failure, revascularisation (AND
DIFFERENT COMPOSITES OF THIS OUTCOME)
Studies will be excluded if they:
 are validation studies
 are prognostic studies that only give one of the following ABPM: day, night, 24hours. Need
to compare all of them for which is best predictor
 assess the prediction of future BP (rather than the specified clinical outcomes)
347
Hypertension (partial update)
Review protocols
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG18, however the studies included within the SR/MA may
have been published pre- and post- 2003); RCTs; non-RCTs: mainly diagnostic, prognostic
and cohort studies, reliability / reproducibility studies
Taking into consideration the advice on prognostic reviews in the NICE guidelines manual,
meta-analysis or GRADE will not be undertaken for prognostic studies as well as for other
non-RCTs.
E.3
Review protocol for initiating and monitoring treatment,
including blood pressure targets




Blood pressure thresholds for initiating treatment
Blood pressure targets for treatment
Monitoring treatment (covered in AMB1 and AMB2 protocol, xxx)
Equivalent thresholds for intervention / treatment targets
Blood pressure thresholds for initiating treatment and blood pressure targets for
treatment
Review
questions
INITIATION: In adults with primary hypertension, at what BP should treatment be
initiated?
OPTIMUM: In adults with primary hypertension, what is the optimum BP that
should be reached for once treatment has been initiated/ targeted for treatment?
Objectives
This aim of this review is to determine the appropriate level (threshold) of BP that a person
must have in order for treatment to be initiated, and to determine the appropriate BP that
treatment should be targeted to in order to gain an appropriate reduction in risk of CV
events.
Criteria
Population (s):
Adults ≥18 years old with suspected primary hypertension (diagnosis) or hypertension
(treatment) who may or may not have pre-existing cardiovascular disease, chronic kidney
disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s) and comparisons:
Home, ABPM and clinic
Outcome(s):
FOR INITIATION
 Risk of developing clinical outcomes (at different BPs; studies will be excluded if they do
not assess results by stratifying into different BP thresholds / values (thresholds or
treatment targets)
 Mortality
 Stroke
 MI
348
Hypertension (partial update)
Review protocols
 Vascular procedures (including both coronary and arotid artery procedures)
 Angina requiring hospitalisation
 Major adverse cardiac and cerebrovascular events (MAACE): fatal and non-fatal MI, fatal
and non-fatal stroke, hospitalised angina, hospitalised heart failure, revascularisation (AND
DIFFERENT COMPOSITES OF THIS OUTCOME)
FOR OPTIMUM
Same outcomes as above plus:
 Number of patients with controlled BP
 Number of patients reaching target
 Final BP values
Studies will be excluded if they:
 do not stratify results by different BP thresholds / values
 compare treated vs. untreated patients (for OPTIMUM)
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG18, however the studies included within the SR/MA may
have been published pre- and post- 2003); RCTs; non-RCTs: mainly diagnostic, prognostic
and cohort studies.
Taking into consideration the advice on prognostic reviews in the NICE guidelines manual,
meta-analysis or GRADE will not be undertaken for prognostic studies as well as for other
non-RCTs.
Equivalent thresholds for intervention / treatment targets
Review
question
ADJUST (AMB3): if used, should ambulatory or home blood pressure readings be
interpreted differently to office measurements? i.e. are different thresholds for
intervention/targets for treatment required, or should adjustment be made to
readings.
Objectives
To establish the adjustment factor required for ambulatory, home and office BP
measurements in order to give the same (in adults ≥18 years old).
Criteria
See ‘review strategy’
Search
Strategy
See ‘review strategy’
Review
Strategy
For this question data from the reviews comparing all three BP measurement methods
(ABPM, home and clinic) were used to determine equivalent thresholds for intervention
and targets for treatment based on outcome measures showing equivalent prognosis.
E.4
Review protocol for pharmacological interventions
 Step one therapy: Angiotensin-converting enzyme inhibitors (ACEi) versus Angiotensin
Receptor Blockers (ARB); Diuretics
 Step two therapy: A+C vs A+D
 Resistant hypertension
 Special groups for consideration: people aged over 80 years and ethnicity
Step-one therapy: ACEi vs ARBs
Review
In adults with primary hypertension, which is the most clinically and cost effective
349
Hypertension (partial update)
Review protocols
question
Objectives
Criteria
anti-hypertensive monotherapy (ACEi vs ARB) for first-line treatment, and does this
vary with age and ethnicity?
Since the publication of the earlier HT guidelines, a major outstanding issue for first line
therapy was A vs A (as there was no evidence for these comparisons). Recent evidence has
now emerged directly comparing ACEi and ARB. Therefore the aim of this review is to
estimate the efficacy, safety and cost-effectiveness of ACEi vs ARB for first-line treatment
of adults ≥18 years old with primary hypertension.
Population (s):
Adults ≥18 years old with primary hypertension who may or may not have pre-existing
cardiovascular disease, chronic kidney disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
ACEi*
Comparison(s):
ARB*
*In accordance with the 2006 guideline, all drugs in these classes will be assessed (licensed
and unlicensed) as we are assuming a class effect
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where reported, silent
MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary and carotid
artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is reported by trials)
 Major adverse cardiac and cerebrovascular events (MAACE):
fatal and non-fatal MI, fatal and non-fatal stroke, hospitalised
angina, hospitalised heart failure, revascularisation (AND
DIFFERENT COMPOSITES OF THIS OUTCOME)
10%
10%
Safety
 Study drug withdrawal rates (surrogate for adverse effects of
drug treatment and for adherence)
 Angioodema in black people of African and Caribbean descent
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG34, however the studies included within the SR/MA may
have been published pre- and post- 2006); RCTs.
Where appropriate, meta-analysis will be undertaken.
Subgroup analyses will be done for age >80 years (vs <80 years) and for black people of
African and Caribbean descent (vs. white people)
350
Hypertension (partial update)
Review protocols
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Studies will be excluded if they:
 have sample size of N<200 (in accordance with the 2006 guideline exclusion criteria)
 have follow-up time of <12 months (in accordance with the 2006 guideline exclusion
criteria)
Step-one therapy: diuretics
Review
question
In adults with primary hypertension, which is the most clinically and cost effective
thiazide diuretic (bendrofluazide / bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide) for first-line treatment, and does this vary with age and
ethnicity?
Objectives
A major issue for first line therapy is which diuretic is most effective because in the UK we
use bendrofluazide for which there is very little evidence and the rest of the world do not
use this drug. Is there a better diuretic to use? This was not looked at in the previous
guidelines and so we have searched all dates. Therefore the aim of this review is to to
estimate the efficacy, safety and cost-effectiveness of specific diuretics (bendrofluazide /
bendroflumethiazide, chlorthalidone, indapamide, hydrochlorothiazide) for first-line
treatment of adults ≥18 years old with primary hypertension.
Criteria
Population (s):
Adults ≥18 years old with primary hypertension who may or may not have pre-existing
cardiovascular disease, chronic kidney disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
TD / TDLs (bendrofluazide / bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide)
Comparison(s):
TD/TDLs (head-to head – using the four named drugs above), placebo, other a-HT drug
classes (BB, CCB, ACEi, ARBs: all drugs – licensed and unlicensed as we are assuming a
class effect)
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where reported,
silent MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary and carotid
artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is reported by
trials)
 Major adverse cardiac and cerebrovascular events
351
10%
Hypertension (partial update)
Review protocols
(MAACE): fatal and non-fatal MI, fatal and non-fatal
stroke, hospitalised angina, hospitalised heart failure,
revascularisation (AND DIFFERENT COMPOSITES OF
THIS OUTCOME)
 BP lowering (for part 2 of the question only – see review
strategy section)
10%
Safety
 Study drug withdrawal rates (surrogate for adverse effects
of drug treatment and for adherence)
 Angioodema in black people of African and Caribbean
descent
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG34, however the studies included within the SR/MA may
have been published pre- and post- 2006); RCTs.
Where appropriate, meta-analysis will be undertaken.
Subgroup analyses will be done for age >80 years (vs <80 years) and for black people of
African and Caribbean descent (vs. white people)
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Review strategy for part 1 and 2 of the question:
PART 1: TDs/TDLs vs placebo or other a-HT drug classes
TDs/TDLs must be: bendrofluazide / bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide
Sample size must be N>200 (in accordance with the 2006 guideline exclusion criteria
Follow up must be ≥1 year (in accordance with the 2006 guideline exclusion criteria
Outcomes must be clinical only (not BP)
Exclude studies if they report AEs / withdrawals but not our pre-specified clinical
outcomes
PART 2: Head-to-head - TDs/TDLs vs other TDs/TDLs
TDs/TDLs must be: bendrofluazide / bendroflumethiazide, chlorthalidone, indapamide,
hydrochlorothiazide
Sample size: any
Follow up: any
Outcomes: BP only (only use clinical if reported ≥1 year results)
Step-two therapy
Review
question
Objectives
Criteria
COMBI: In adults with primary hypertension, which is the most clinically and cost
effective combination of anti-hypertensives (A+C or A+D) for second line treatment,
and does this vary with age and ethnicity?
The earlier HT guideline recommended the use of adding in either CCBs or D for secondline therapy (ie. A+C or A+D). Recent evidence has now emerged directly comparing these
combinations. The aim of this review is therefore to estimate the efficacy, safety and costeffectiveness of ACEi + CCB or ARB + CCB vs. ACE + Diuretic or ARB + Diuretic) for
second-line treatment of adults ≥18 years old with primary hypertension.
Population (s):
Adults ≥18 years old with primary hypertension who may or may not have pre-existing
cardiovascular disease, chronic kidney disease and diabetes
352
Hypertension (partial update)
Review protocols
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
A (ACEi or ARB)* + CCB* in second line treatment
Comparison(s):
A (ACEi or ARB)* + D* in second line treatment
*In accordance with the 2006 guideline, all drugs in these classes will be assessed (licensed
and unlicensed) as we are assuming a class effect
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where reported,
silent MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary and
carotid artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is reported by
trials)
 Major adverse cardiac and cerebrovascular events
(MAACE): fatal and non-fatal MI, fatal and non-fatal
stroke, hospitalised angina, hospitalised heart failure,
revascularisation (AND DIFFERENT COMPOSITES
OF THIS OUTCOME)
10%
10%
Safety
 Study drug withdrawal rates (surrogate for adverse
effects of drug treatment and for adherence)
 Angioodema in black people
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG18, however the studies included within the SR/MA may
have been published pre- and post- 2003); RCTs.
Where appropriate, meta-analysis will be undertaken.
Subgroup analyses will be done for age >80 years (vs <80 years) and for black people of
African and Caribbean descent (vs. white people)
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Studies will be excluded if they:
 have sample size of N<200 (in accordance with the 2004 guideline exclusion criteria)
 have follow-up time of <12 months (in accordance with the 2004 guideline exclusion
criteria)
353
Hypertension (partial update)
Review protocols
Resistant hypertension
Review
question
Objectives
Criteria
In adults with resistant hypertension, which is the most clinically and cost effective
fourth-line pharmacological treatment, and does this vary with age and ethnicity?
Resistant hypertension has not been covered in the previous guidelines. The aim of this
review is therefore to estimate the efficacy, safety and cost-effectiveness of fourth-line
therapy in adults ≥18 years old with resistant hypertension.
Population (s):
Adults (≥18 years old) with resistant hypertension (people whose BP is still uncontrolled
despite treatment with optimal doses of three a-HT drugs) who may or may not have preexisting cardiovascular disease, chronic kidney disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
4th line drugs (MOST WIDELY USED OPTIONS: including alpha-blockers; betablockers; other/further diuretics such as amiloride and spironolactone; aliskerin; aldosterin
antagonists; moxonidine )
Comparison(s):
Any comparison (placebo or each other)
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where
reported, silent MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary
and carotid artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is
reported by trials)
 Major adverse cardiac and cerebrovascular
events (MAACE): fatal and non-fatal MI, fatal
and non-fatal stroke, hospitalised angina,
hospitalised heart failure, revascularisation
(AND DIFFERENT COMPOSITES OF THIS
OUTCOME)
10%
10%
Safety
 Study drug withdrawal rates (surrogate for
adverse effects of drug treatment and for
adherence)
 Angioodema in black people
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG18, however the studies included within the SR/MA may
354
Hypertension (partial update)
Review protocols
have been published pre- and post- 2003); RCTs; cohort studies.
Where appropriate, meta-analysis will be undertaken.
Subgroup analyses will be done for age >80 years (vs <80 years) and for black people of
African and Caribbean descent (vs. white people)
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Studies will be excluded if they:
 are RCTs with a sample size of N<200 (in accordance with the 2004 guideline exclusion
criteria); unless evidence is sparse
 are RCTs with a follow-up time of <12 months (in accordance with the 2004 guideline
exclusion criteria); unless evidence is sparse
355
Hypertension (partial update)
Review protocols
Special groups for consideration: people aged over 80 years
Review
question
In adults with primary hypertension, what is the most clinically and cost effective
first-line anti-hypertensive treatment (drug classes) in elderly people (aged ≥80
years)?
Objectives
In the previous guidelines (CG18 and CG34) there was no evidence available showing
whether there were differences in treatment effects in elderly people (80+). However data
has since then emerged on drug treatment in this age-group. Therefore the aim of this
review is to to estimate the efficacy, safety and cost-effectiveness of anti-hypertensive
drugs for the first-line treatment of adults ≥80 years old with primary hypertension.
Criteria
Population (s):
Adults ≥80 years old with primary hypertension who may or may not have pre-existing
cardiovascular disease, chronic kidney disease and diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
All a-HT drug classes (BB, CCB, ACEi, ARBs)*
Comparison(s): all a-HT drug classes (BB, CCB, ACEi, ARBs)*
*In accordance with the 2006 guideline, all drugs in these classes will be assessed (licensed
and unlicensed) as we are assuming a class effect
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where
reported, silent MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary
and carotid artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is
reported by trials)
 Major adverse cardiac and cerebrovascular
events (MAACE): fatal and non-fatal MI,
fatal and non-fatal stroke, hospitalised
angina, hospitalised heart failure,
revascularisation (AND DIFFERENT
COMPOSITES OF THIS OUTCOME)
10%
10%
Safety
 Study drug withdrawal rates (surrogate for
adverse effects of drug treatment and for
adherence)
 Angioodema in black people of African and
Caribbean descent
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG34, however the studies included within the SR/MA may
356
Hypertension (partial update)
Review protocols
have been published pre- and post- 2006); RCTs.
Where appropriate, meta-analysis will be undertaken.
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Studies will be excluded if they:
 have sample size of N<200 (in accordance with the 2006 guideline exclusion criteria)
 have follow-up time of <12 months (in accordance with the 2006 guideline exclusion
criteria)
Special groups for consideration: ethnicity
Review
question
Objectives
Criteria
In adults with primary hypertension, what is the most clinically and cost effective
first-line anti-hypertensive treatment (drug classes) in black people of African or
Caribbean descent)?
In the previous guideline there was little evidence available showing whether there were
differences in treatment effects in elderly people (80+). However data has since then
emerged on drug treatment in this age-group. Therefore the aim of this review is to to
estimate the efficacy, safety and cost-effectiveness of anti-hypertensive drugs for the firstline treatment of adults ≥80 years old with primary hypertension.
Population (s):
Black adults (of African or Caribbean descent) ≥18 years old with primary hypertension
who may or may not have pre-existing cardiovascular disease, chronic kidney disease and
diabetes
Studies in indirect populations will not be considered (ocular HT, pulmonary HT, HT
during pregnancy, acute HT, malignant HT, portal HT, renal HT and intercranial HT)
Studies in adult women of conception age will be considered (in accordance with BNF).
Studies with an exclusive diabetic or CKD population will be excluded (in accordance with
the 2006 guideline exclusion criteria)
Intervention (s):
ACEi*
Comparison(s):
ARB*
*In accordance with the 2006 guideline, all drugs in these classes will be assessed (licensed
and unlicensed) as we are assuming a class effect
Outcome(s):
Effectiveness
 Mortality from any cause
10%
 Stroke (ischaemic or haemorrhagic)
 Myocardial infarction (MI) (including, where
reported, silent MI)
 Heart failure
 New onset diabetes
 Vascular procedures (including both coronary and
carotid artery procedures)
 Angina requiring hospitalisation
 Health-related quality of life (to use what is
reported by trials)
 Major adverse cardiac and cerebrovascular events
357
10%
Hypertension (partial update)
Review protocols
(MAACE): fatal and non-fatal MI, fatal and nonfatal stroke, hospitalised angina, hospitalised heart
failure, revascularisation (AND DIFFERENT
COMPOSITES OF THIS OUTCOME)
10%
Safety
 Study drug withdrawal rates (surrogate for adverse
effects of drug treatment and for adherence)
 Angioodema
Search
Strategy
See appendix Appendix D:Search Strategies.
Review
Strategy
Study design: SRs / meta-analyses (will be included if they are published after the cut-off
date of the previous guideline CG34, however the studies included within the SR/MA may
have been published pre- and post- 2006); RCTs; sub-group analyses of RCTs, cohort
studies.
Where appropriate, meta-analysis will be undertaken.
Sensitivity analysis will be carried out based on methodological quality if significant
heterogeneity exists.
Overall assessment of the quality (for each outcome) will be undertaken using GRADE.
Studies will be excluded if they have a:
 sample size of N<1000 in each arm (as a very large study, ALLHAT, has been published
with a sample size of 42,418)
 follow-up time of <12 months (in accordance with the 2006 guideline exclusion criteria)
E.4.1
Health economic review protocol
Review
question
All questions – health economic evidence
Objectives
To identify economic studies relevant to the review questions set out above.
Criteria
Populations, interventions and comparators, and date cut-offs as specified in the questionspecific review protocols. Must be a relevant economic study design (cost-utility analysis,
cost-benefit analysis, cost-effectiveness analysis, cost-consequence analysis, comparative
cost analysis).
Search strategy
See Appendix D:.
Review
strategy
Each study is assessed using the NICE economic evaluation checklist – NICE (2009)
Guidelines Manual, Appendix H.
Inclusion/exclusion criteria
 If a study is rated as both ‘Directly applicable’ and ‘Minor limitations’ (using the NICE
economic evaluation checklist) then it should be included in the guideline. An evidence
table should be completed and it should be included in the economic profile.
 If a study is rated as either ‘Not applicable’ or ‘Very serious limitations’ then it should
be excluded from the guideline. It should not be included in the economic profile and
there is no need to include an evidence table.
 If a study is rated as ‘Partially applicable’ and/or ‘Potentially serious limitations’ then
there is discretion over whether it should be included. The health economist should
make a decision based on the relative applicability and quality of the available evidence
for that question, in discussion with the GDG if required. The ultimate aim being to
include studies that are helpful for decision making in the context of the guideline.
Where exclusions occur on this basis, this should be noted in the relevant section of the
guideline with references.
Also exclude:
 unpublished reports unless submitted as part of the call for evidence
358
Hypertension (partial update)
Review protocols
Review
question
All questions – health economic evidence
 abstract-only studies
 letters
 editorials
 reviews of economic evaluations(a)
 foreign language articles
Where there is discretion
The health economist should be guided by the following hierarchies.
Setting:
1. UK NHS
2. OECD countries with predominantly public health insurance systems (e.g. France,
Germany, Sweden)
3. OECD countries with predominantly private health insurance systems (e.g. USA,
Switzerland)
4. Non-OECD settings (always ‘Not applicable’)
Economic study type:
1. Cost-utility analysis
2. Other type of full economic evaluation (cost-benefit analysis, cost-effectiveness
analysis, cost-consequence analysis)
3. Comparative cost analysis
4. Non-comparative cost analyses including cost of illness studies (always ‘Not
applicable’)
Year of analysis:
 The more recent the study, the more applicable it is
Quality and relevance of effectiveness data used in the economic analysis:
 The more closely the effectiveness data used in the economic analysis matches with the
studies included for the clinical review the more useful the analysis will be to decision
making for the guideline.
(a) Recent reviews will be ordered although not reviewed. The bibliographies will be checked for relevant studies, which will
then be ordered.
359
F.1
Blood pressure variability
STUDY 1
360
P. M.
Rothwell, S.
C. Howard,
E. Dolan, E.
O'Brien, J.
E. Dobson,
B. Dahlof,
N. R.
Poulter, and
P. S. Sever.
Effects of
beta blockers
and calciumchannel
blockers on
withinindividual
variability in
blood
pressure and
risk of
stroke.
Lancet
Neurol 9
(5):469-480,
2010.
1) RCT Anglo-Scandinavian
Cardiac Outcomes Trial Blood
Pressure Lowering Arm (ASCOTBPLA)
Country not stated
Selection, randomisation,
allocation concealment, power
calculation, attrition not described
(referenced to another paper)
Participants in two treatment
groups well matched
Intention to treat analysis
Medications added in to patients in
either group
Interpretation of apparent
correlates with treatment effects in
trials on the basis of data collected
after randomisation potentially
subject to bias
No differences between groups in
loss to follow up
2) Medical Research Council trial
ID 15883
Country not stated
1) whole
study
19257;
used in
this
analysis
18530
(96.2%)
who had
at least 2
scheduled
follow up
visits
from 6
months
onwards
2) 4396
1) Inclusion: Patients
with hypertension aged
40-79 years and had at
least 3 other
cardiovascular risk
factors but no previous
history of coronary heart
disease. Clinic BPs
(CBP) 6-monthly using
validated semi-automatic
oscillometric device,
seated and rested for 5
minutes; measured 3
times at 5 minute
intervals and mean of
last 2 measurements
used. Participants at 4
centres had repeated
annual ambulatory BPs
using validated monitors
giving daytime (09002100) and night-time
(0100-0600) and 24-hour
readings.
2) Hypertensive patients
(mean SBP 160209mmHg and mean
DBP <115mmHg) aged
1) amlodipineperindopril to
achieve CBP
target
≤140/90mmHg
without diabetes
or ≤130/80mmHg
with diabetes,
plus other
antihypertensives
as needed
1) atenololthiazide to
achieve CBP
target
≤140/90mmHg
without diabetes
or ≤130/80mmHg
with diabetes
plus other
antihypertensives
as needed
2) atenolol 50mg
daily or a diuretic
combination
(hydrochlorothiaz
ide 25mg plus
amiloride 2.5mg
daily), titrated to
achieve clinic
SBP <150mmHg
(if mean run-in
SBP 160179mmHg) or
<160mmHg (if
mean run-in SBP
≥180mmHg)
2) placebo
1)
Median
5.5
years
Primary
endpoint: 1)
Risk of
stroke and
coronary
events
2) Stroke
risk
Visit-tovisit
variability
in BP was a
strong
predictor of
long-term
risk of
stroke; this
was
increased by
atenolol
none
Hypertension (partial update)
Clinical evidence tables
Appendix F: Clinical evidence tables
65-74 years. Clinic BP
measured 3 times sitting
using random zero
sphygmomanometer;
mean of 2nd 2 readings
used.
No data for individual patients on
add-on treatments with other
drugs; use of other drugs was
particularly high in atenolol group
(52% vs. 38% in diuretic group at
5 years) with drugs (e.g.
nifedipine) that would have
reduced variability
Baseline characteristics:
1) ASCOT-BPLA:
361
Systolic BP:
Mean
Maximum
Minimum
Within-individual visit-to-visit variability:
Standard deviation
Coefficient of variation (SD/mean)
Within-visit variability:
Standard deviation
Range of 3 readings
Diastolic BP:
Mean
Maximum
Minimum
Within-individual visit-to-visit variability:
Standard deviation
Coefficient of variation (SD/mean)
2) MRC trial:
Atenolol-based regimen
Amlodipine-based regimen
Difference
141.8 (13.0)
164.2 (18.9)
122.6 (13.5)
139.1 (11.1)
157.4 (16.1)
123.0 (11.8)
2.68 (2.58-2.78), p<0.0001
6.80 (6.68-6.92), p<0.0001
-0.40 (-0.50 to +0.30)
13.42 (5.77)
9.41 (3.78)
10.99 (4.79)
7.87 (3.23)
2.43 (2.36-2.50), p<0.0001
1.54 (1.49-1.59)
5.91 (0.02)
5.16 (0.04)
5.42 (0.02)
4.85 (0.04)
0.49 (0.44-0.54)
0.31 (0.20-0.42)
82.1 (7.6)
93.5 (9.6)
71.8 (8.4)
80.2 (7.4)
90.4 (9.0)
70.8 (8.1)
1.98 (1.90-2.06)
3.10 (3.00-3.20)
1.00 (0.90 to 1.10)
6.98 (2.72)
8.54 (3.30)
6.26 (2.42)
7.86 (3.04)
0.72 (0.67-0.77)
0.68 (0.63-0.73)
Hypertension (partial update)
Clinical evidence tables
STUDY 1
2) MRC trial:
362
Atenolol increased visit-to-visit variability in SBP compared to placebo, whereas the diuretic combination did not. Mean SBP was higher in atenolol group than diuretic
group (156.6 (12.1) mmHg vs. 151.2 (12.1) mmHg) in the first 18 months of the trial, mainly due to a higher maximum SBP (178.2 (16.1)mmHg vs. 168.8 (15.7)mmHg)
as a consequence of increased variability, with little difference in minimum SBP (135.9 (13.5)mmHg vs. 134.2 (12.7)mmHg). The risk of stroke in the atenolol group was
higher than placebo for the first 2 years (HR 1.31, 0.81-2.10) when variability was also higher than with placebo (SD 14.38 (5.34) vs. 12.12 (4.48), p<0.001) despite
substantially lower mean BP (156.6 (12.1)mmHg vs. 167.4 (12.0)mmHg, p<0.001). The risk of stroke was lower on atenolol than placebo after 2 years of follow up (HR
0.62, 0.40-0.94), by which time the difference in mean BP was still large 151.8 (13.0)mmHg vs. 166.3 (14.2)mmHg, p<0.001) but variability no longer differed (SD
12.30 (6.24) vs. 12.68 (6.21), p=0.12).
STUDY 2
P. M.
Rothwell, S.
C. Howard,
E. Dolan, E.
O'Brien, J.
E. Dobson,
B. Dahlof, P.
S. Sever, and
N. R.
Poulter.
Prognostic
significance
1a) RCT UK-TIA
aspirin trial and 3
similar validation
cohorts:
1b) RCT European
Stroke Prevention
Study (ESPS-1)
placebo group only
1c) RCT Dutch TIA
trial
1d) Subgroup of RCT
1a) total
2435 with
Recent TIA
or
ischaemic
stroke; this
analysis:
2006 TIA
only
1b) 2500
1c) 3150
1a) Recent TIA Sitting BP
measured once every 4 months
with mercury
sphygmomanometer and patient
rested. 4 categories: stable
normotension (maximum
≤140mmHg); episodic moderate
hypertension (minimum
≤140mmHg and maximum 140179mmHg); episodic severe
hypertension (minimum
≤140mmHg and maximum
1a) Aspirin
1200mg or
300mg
1b) dipyridamole
75mg + aspirin
325mg three
times daily
1c) aspirin 283mg
or atenolol 50mg
1d and 2)
1a) Placebo
1b) Placebo
1c) aspirin 30mg
or placebo
1d and 2)
atenolol-thiazide
to achieve CBP
target
≤140/90mmHg
without diabetes
1a)
median
10
(range
1-20)
follow
ups at 4monthly
intervals
1b) 2
years
1a) Vascular
events and
deaths
1b-d) Stroke
2) Risk of
stroke and total
coronary
events
Visit-to-visit
variability in
none
Hypertension (partial update)
Outcomes:
1) ASCOT-BPLA:
Hazard ratio for stroke with randomised treatment (amlodipine versus atenolol, 0.78, 0.67-0.90, p=0.001) diminished less after adjustment for mean SBP during follow up
(0.84, 0.72-0.98, p=0.025) than after adjustment for visit-to-visit variability using standard deviation of SBP (0.94, 0.81-1.10, p=0.47). Similarly for risk of coronary
events (0.85, 0.77-0.94, p=0.002 changed to 0.88, 0.80-0.98, p=0.019 after adjusting for mean SBP and to 1.00, 0.90-1.11, p=0.96 after adjusting for SD of SBP).
Adjustment for variability in DBP had less effect. Group SD of SBP decreased on treatment with amlodipine and increased in the atenolol group, independent of effects
on mean BP. The reduced event rates in the amlodipine group could not be fully accounted for by changes in mean BP or other risk factors but were explained by
variability. Patients with good control of BP but high residual variability in SBP had a 5 times higher risk of stroke than those with low residual SBP variability.
Clinical evidence tables
STUDY 1
No treatment-group differences in group SD and coefficient of variation of SBP at baseline.
363
Variability in BP
measurements could
be due to nonadherence to
guidelines for
measurement or
inadequate calibration
of measuring devices
BP measured only
once in 1a and 1c
No data on use of, or
compliance with,
antihypertensive drugs
during follow up in
older TIA cohorts
Findings cannot be
generalised to healthy
cohorts
2) RCT AngloScandinavian Cardiac
Outcomes Trial Blood
Pressure Lowering
Arm (ASCOT-BPLA)
1d) 2011
2) 18530
(96% of
total in
study) had
2 or more
visits from
6 months
onwards
(median 10,
IQR 9-11)
≥180mmHg); and stable
hypertension (minimum
>140mmHg)
1b) mean of left and right arm
sitting after rest, mercury
sphygmomanometer every 3
months
1c) Sitting BP measured once
every 4 months with mercury
sphygmomanometer and patient
rested
1d) Previous TIA or stroke
2) Patients with hypertension
aged 40-79 years and had at
least 3 other cardiovascular risk
factors but no previous history
of coronary heart disease. Clinic
BPs (CBP) 6-monthly using
validated semi-automatic
oscillometric device, seated and
rested for 5 minutes; measured
3 times at 5 minute intervals and
mean of last 2 measurements
used. Participants at 4 centres
had repeated annual ambulatory
BPs using validated monitors
giving daytime (0900-2100) and
night-time (0100-0600) and 24hour readings.
For all: Visit-to-visit variability
defined as Standard deviation
(SD) and Coefficient of
variation (SD/mean) or SD and
range of 3 readings at 1 visit
amlodipineperindopril to
achieve CBP
target
≤140/90mmHg
without diabetes
or ≤130/80mmHg
with diabetes,
plus other
antihypertensives
as needed
or ≤130/80mmHg
with diabetes plus
other
antihypertensives
as needed
1c)
mean
2.6
years
2)
Median
5 years
SBP was a
strong
predictor of
subsequent
stroke
independent of
mean SBP;
maximum SBP
reached was
also a strong
predictor of
subsequent
stroke
independent of
mean SBP.
Increased
residual
variability in
SBP in patients
with treated
hypertension is
associated with
a high risk of
vascular
events.
Hypertension (partial update)
ID 6157
Anglo-Scandinavian
Cardiac Outcomes
Trial Blood Pressure
Lowering Arm
(ASCOT-BPLA)
Clinical evidence tables
STUDY 1
of visit-tovisit
variability,
maximum
systolic
blood
pressure, and
episodic
hypertension
. Lancet 375
(9718):895905, 2010.
Covariates: age, gender,
baseline vascular risk factors
Rothwell 6157
Baseline characteristics:
1a-d) UK-TIA aspirin trial: 2006 patients, 1438 men, mean age 60.3 (9.1) years; age and gender not specified for other cohorts. Available data:
ASCOT-BPLA subgroup with previous TIA
364
UK-TIA aspirin trial
Atenolol group
Amlodipine group
ESPS-1 placebo group
Dutch TIA trial
n
1324
1012
999
1247
3150
Mean baseline SBP (mmHg)
150.2 (25.3)
163.7 (18.7)
164.4 (17.9)
156.3 (22.7)
157.9 (26.3)
Mean 1 year SBP (mmHg)
146.6 (23.4)
148.3 (19.7)
143.3 (17.4)
154.8 (22.3)
151.7 (22.5)
Mean Visit-to-visit variability:
SD
Coefficient of variation (SD/mean)
14.2 (6.6)
9.6 (3.9)
14.4 (6.1)
10.00 (4.0)
11.4 (4.9)
8.2 (3.3)
14.6 (6.8)
9.3 (4.1)
14.9 (6.4)
9.7 (3.9)
Outcomes:
1a) UK-TIA aspirin trial: 1324 (66%) patients reached visit 7, of whom 270 had a stroke or coronary event. Mean SBP over visits 1-7 predicted stroke: adjusted HR 1.43
(1.18-1.74) per 20mmHg, p<0.0001. Visit-to-visit variability in SBP was a stronger predictor: adjusted HR 4.37 (2.73-6.99) independent of mean SBP in patients
receiving (HR 3.67, 2.34-5.75) and not receiving (HR 2.27, 1.41-3.67) antihypertensives at baseline. The effect was similar for men and women but decreased with age:
9.43 (1.96-45.5) at <56 years; 3.01 (0.97-9.36) at 56-64 years and 1.71 (0.74-3.98) at ≥65 years. Maximum SBP was more predictive than mean SBP, and was most
predictive at lower values of mean SBP: HR for maximum SBP adjusted for mean SBP: 4.95 (1.28-22.4, p=0.007) at mean SBP <130mmHg; 3.19 (1.65-6.23, p=0.0001)
at 130-159mmHg and 1.13 (0.50-2.53, p=0.75) at ≥160mmHg. Patients with episodic severe hypertension (minimum ≤140mmHg and maximum ≥180mmHg) had a
higher risk of stroke than those with stable hypertension (13.7% vs. 4.5%, p=0.003; age and gender-adjusted HR 3.58, 1.58-8.10) despite a lower mean SBP (157.9
(8.7)mmHg vs. 167.3 (7.2)mmHg, p=0.001.
1a-d): Hazard ratios for stroke (top versus bottom decile of each measure):
ASCOT-BPLA subgroup with previous TIA
UK-TIA aspirin trial
Atenolol group
Amlodipine group
ESPS-1 placebo group
Dutch TIA trial
Hypertension (partial update)
Clinical evidence tables
STUDY 1
3.63 (2.41-5.48)
1.81 (0.89-3.67)
0.94 (0.36-2.42)
1.89 (0.96-3.71)
2.34 (1.41-3.89)
SD SBP adjusted for mean SBP
4.84 (3.03-7.74)
4.29 (1.78-10.36)
4.39 (1.68-11.50)
1.78 (1.21-2.62)
3.35 (1.63-6.87)
Coefficient of variation SBP
adjusted for mean SBP
3.82 (2.54-5.73)
3.51 (1.56-7.93)
3.25 (1.32-8.00)
2.22 (1.52-3.22)
3.41 (1.62-7.19)
2) Patients with episodic severe hypertension (minimum ≤140mmHg and maximum ≥180mmHg) had a higher risk of stroke than those with stable hypertension (4.0% vs.
2.7%, p=0.03 despite a lower mean SBP (142.1 (14.8)mmHg vs. 147.3 (13.8)mmHg, p<0.0001). Maximum SBP predicted stroke (e.g. top decile Hazard ratio for risk of
stroke in atenolol group 2.51, 1.69-3.73, p=0.0008).
Variability was related to factors that correlate with arterial stiffness including age, female gender, smoking, diabetes and peripheral vascular disease (data not given).
STUDY 3
365
A. J. Webb,
U. Fischer,
Z. Mehta,
and P. M.
Rothwell.
Effects of
antihyperten
sive-drug
class on
interindividu
al variation
in blood
pressure and
risk of
stroke: a
systematic
review and
metaanalysis.
Lancet 375
(9718):906-
Meta-analysis
Country not applicable
From published systematic
reviews from Medline and
Cochrane databases, reference
lists of reviews, no language
restrictions
Not individual patient data, so
not possible to exclude from
analysis people who had a
non-fatal event during early
follow up but whose BP data
contributed to group means
thereafter (but potential bias
assessed as small by authors)
Low rate of reporting of BP
variability in trials (assessed
as unlikely to introduce bias
389
RCTs for
outcome
of BP
variabilit
y; 21 for
clinical
outcomes
DRUGS
vs
DRUGS
Inclusion
criteria
referenced to
web appendix:
briefly, trials of
the drug classes
listed
Exclusion: no
valid
comparison
group, lasted <2
weeks,
ineligible
patient group,
BP or group
variation in BP
not reported
Covariates: age,
gender, ethnic
8 main drug
classes: thiazide
and thiazide-like
diuretics;
β blockers;
ACE inhibitors;
angiotensin-2receptor blockers;
dihydropyridine
calcium channel
blockers; nondihydropyridine
calcium channel
blockers; α1
blockers; placebo
1) versus
all of the
other of
the 8
drug
classes or
placebo
2) versus
each of
the other
of the 8
drug
classes or
placebo
one at a
time
2 weeks
for BP
variability
; 1 year
for
clinical
outcomes
For trials with >100 patients
per treatment group with at
least 1 year of follow up: risk
of stroke, MI, heart failure,
cardiovascular mortality
Calcium channel blockers,
non-dihydropyridine calcium
channel blockers and diuretics
reduced variability, whereas
angiotensin-receptor blockers,
ACE inhibitors and β blockers
increased it. Effects of
treatment on variation in SBP
were correlated with effects on
risk of stroke (but not MI or
heart failure) independently of
differences in mean SBP.
none
Hypertension (partial update)
Mean SBP (unadjusted)
Clinical evidence tables
STUDY 1
origin, diabetes,
renal failure
ID 15886
Webb 15886
Baseline characteristics:
Changes in inter-individual variance (SD2) in BP (a surrogate for within-individual variability) expressed as the ratio of the variances follow up versus baseline (VR) and
as the percentage differences in coefficient of variation follow up versus baseline (CV) both within trial (between treatment groups) and across trials (for each drug class).
SD and CV correlated with each other but not with mean SBP. Significant heterogeneity among the 682 treatment groups (p<0.001), 24.7% of which explained by drug
class after adjustment for age, gender, ethnic origin, diabetes, renal failure, baseline SBP and DBP.
SD of SBP at follow up and variance ratio
366
Patients
Trials
SD (95% CI)
Calcium channel blockers
34221
134
14.97 (14.74-15.46)
Ratio of variances follow up versus baseline: VR (95%
CI)
0.89 (0.82-0.97)
Non-dihydropyridine calcium channel blockers
6777
33
15.92 (15.31-16.76)
0.96 (0.84-1.10)
Non-loop diuretics
30090
58
15.65 (15.24-16.17)
1.17 (1.01-1.35)
Angiotensin-2-receptor blockers
20748
61
16.74 (16.35-17.07)
1.20 (1.04-1.39)
ACE inhibitors
19235
160
16.95 (16.49-17.43)
1.00 (0.91-1.10)
β blockers
20255
96
17.25 (16.83-17.70)
1.15 (1.02-1.28)
α1 blockers
9540
13
17.00 (16.52-17.64)
1.33 (1.08-1.63)
Placebo
14514
119
17.29 (16.79-17.75)
1.26 (1.09-1.47)
Outcomes:
Within trial comparisons between drug classes on inter-individual variability of SBP at follow up:
Drug class versus placebo
Drug class versus all other drug classes (or placebo)
Patients
Patients
Trials
Ratio of variances follow up versus
baseline: VR (95% CI), p value
Trials
Ratio of variances follow up versus
baseline: VR (95% CI), p value
Hypertension (partial update)
on the basis of funnel plots
referenced to web appendix)
Clinical evidence tables
STUDY 3
915, 2010.
11294
34
0.76 (0.67-0.85), p<0.0001
106697
94
0.81 (0.76-0.86), p<0.0001
Non-dihydropyridine calcium
channel blockers
736
6
0.76 (0.51-1.12), p=0.17
2753
39
0.81 (0.66-0.98), p=0.035
Non-loop diuretics
12187
20
0.91 (0.80-1.03), p=0.15
81772
69
0.87 (0.79-0.96), p=0.007
Angiotensin-receptor
blockers
6756
13
0.93 (0.87-1.00), p=0.044
39447
47
1.16 (1.07-1.25), p=0.0002
ACE inhibitors
7074
43
0.94 (0.82-1.08), p=0.39
76064
125
1.08 (1.02-1.15), p=0.008
β blockers
1210
19
1.04 (0.88-1.23), p=0.65
39392
78
1.17 (1.07-1.28), p=0.0007
Calcium channel blockers, non-dihydropyridine calcium channel blockers and diuretics reduced variability, whereas angiotensin-receptor blockers, ACE inhibitors and β
blockers increased it.
Clinical outcomes: odds ratio (OR) of outcome during follow up in treatment group with lower SD of SBP (lower variability):
367
Risk of stroke
Risk of MI
Risk of heart failure
VR <1.00
OR 0.87 (0.77-0.97), p=0.12
1.01 (0.91-1.08), p=0.45
NS
VR ≤0.80
OR 0.79 (0.71-0.87), p<0.0001
not stated
not stated
Calcium channel blockers vs. all other drugs
OR 0.88 (0.83-0.94), p=0.0002
NS
not stated
β blockers vs. all other drugs
OR 1.19 (1.01-1.42), p=0.0394
NS
not stated
In a model including mean SBP and VR SBP, each variable was independently related to stroke risk (mean SBP p=0.038; VR p=0.014).
Hypertension (partial update)
Calcium channel blockers
Clinical evidence tables
STUDY 3
Diagnosis of hypertension
L. R. Krakoff. Cost-effectiveness of ambulatory blood pressure: a reanalysis. Hypertension 47 (1):29-34, 2006.
Study details
Population &
interventions
Economic analysis: Cost analysis
Population:
People identified as
hypertensive based
on CBPM
3
Study design: Decision analytic model
Approach to modelling: Model
calculates the different numbers of
treated patients taking into account
prevalence of white coat hypertension,
annual incidence of new true
hypertension in those with white coat
hypertension and annual loss to followup and treatment.
6
8
Perspective: USA, healthcare payer
Time horizon: 5 years
Discounting: Costs: none; Outcomes:
n/a
Health
outcomes
None
Costs
Total costs (mean per patient):
Intvn 1: £984
Cost
effectiveness
None
Intervention 1:
No further tests to
confirm diagnosis,
annual follow-up
with CBPM
Intervention 2:
ABPM to confirm
diagnosis and at
annual follow-up
WCH = baseline prevalence of white coat hypertension; Conv = annual
conversion rate of white coat hypertension to true hypertension
Currency & cost year: USA dollars, cost year unclear assumed to be
2005† (presented here as 2005 UK pounds‡)
Cost components incorporated: ABPM; hypertension treatment
(visits, diagnostic tests, pharmacotherapy).
Data sources
Health outcomes: Prevalence white coat hypertension: source not stated but rate varied in analysis 15%-20%. Incidence of new hypertension in those with white coat
hypertension: based on review of literature and varied in analysis. Annual loss to follow-up and treatment: assumed 5% (limited evidence but considered conservative).
Hypertension (partial update)
G.1
Evidence tables – health economic studies (2011 update)
Appendix G: Evidence tables – health economic studies (2011 update)
Quality-of-life weights: n/a. Cost sources: Cost per ABPM use: Centre for Medicare and Medicaid Services of the United States (£47). Annual treatment cost: estimated
based on 5-year survey from US MCO and report that if followed guidelines for using diuretics would reduce by 40% (£212).
Overall quality*: Potentially serious limitations
Overall applicability**: Partially applicable
Abbreviations: ABPM = ambulatory blood pressure monitoring; CBPM = clinic blood pressure monitoring; CI = confidence interval; NR = not reported; MCO = managed care organisation;
NR = not reported.
† Year paper accepted for publication; ‡ Converted using 2005 Purchasing Power Parities468
*Very serious limitations / Potentially serious Limitations / Minor limitations; ** Directly applicable / Partially applicable / Not applicable
G.2
G.3
Initiating and monitoring treatment, including blood pressure targets
G.3.1
Monitoring treatment effect
3
J. A. Staessen, Hond E. Den, H. Celis, R. Fagard, L. Keary, G, and E. T. O'Brien. Antihypertensive treatment based on blood pressure measurement at = home
or in the physician's office: a randomized controlled trial. Journal of the American Medical Association 291 (8):955-964, 2004.
6
9
Study details
Population & interventions
Health outcomes
Costs
Cost effectiveness
Economic analysis:
CCA
Population:
Patients with hypertension (office DBP >95mmHg)
 N = 400
 Mean age = 53 years
 M = 48%
 Belgian 93.2%; Irish 6.8%
Blood pressure:
BP was significantly lower in the
office than the home group.
Total costs (mean per
patient):
Intvn 1: £2811
Intvn 2: £2555
Incremental (2-1):-£256
(CI:NR, p=0.04 )
Basecase ICER (Intvn 2
vs Intvn 1):
N/a. Costs were lower in
the HBPM group but
outcomes blood pressure
control was worse.
Other:
n/a
Subgroup analyses:
None
Study design:
Within RCT analysis
Perspective:
Belgium health
insurance system
Follow-up: 1 year
Discounting: Costs:
n/a; Outcomes: n/a
Intervention 1: Monitoring and treatment
adjustment based on CBPM (average of 3). N=203
Intervention 2: Monitoring and treatment
adjustment based on average of HBPM (average of
6 per day over 1 week). N=197
Antihypertensive medications:
Significantly more home than
office group patients
discontinued antihypertensive
drug treatment.
Left ventricular mass and
reported symptoms:
No significant differences.
Currency & cost year:
2002 Belgium Euros
(presented here as 2002
UK pounds†)
Cost components
Analysis of uncertainty:
n/a
Hypertension (partial update)
Source of funding: NR. Limitations: Does not incorporate all relevant comparators. Does not incorporate health effects (possibly conservative towards ABPM). Some
uncertainty about the applicability of USA costs. Discounting not applied. Source of prevalence of WCH unclear but varied in sensitivity analysis. Limited sensitivity
analysis. Other: none
Evidence tables – health economic studies (2011 update)
Comments
Health outcomes: Within RCT analysis; Quality-of-life weights: n/a. Cost sources: Resource use from within RCT; Belgian units costs from health insurance system
for drugs and physician costs. HBPM from manufacturer.
Comments
Source of funding: AstraZeneca and Pfizer
Limitations: Some uncertainty about applicability of Belgian resource use and unit costs. QALYs not used (cost consequence analysis). In terms of health effects looks at
BP, hypertensive drug use, and LV mass/symptom. Given that blood pressure was significantly different other clinical events and costs of these may be relevant and time
horizon may be insufficient. Within trial analysis and so does not incorporate all available evidence on differences between options (need to confirm if study included in
clinical review and if other evidence exists.
Other: Physician fees and drug costs significantly lower for home monitoring but partially offset by increased cost of home monitoring.
Overall quality*: Potentially serious limitations
Overall applicability**: Partially applicable
Abbreviations: BP = blood pressure; CBPM = clinic blood pressure monitoring; CCA = cost-consequence analysis; CI = confidence interval; DBP = diastolic blood pressure; HBPM = home
blood pressure monitoring; ICER = incremental cost-effectiveness ratio; NR = not reported; RCT = randomised clinical trial.
† Converted using 2002 Purchasing Power Parities468.
*Very serious limitations / Potentially serious Limitations / Minor limitations; ** Directly applicable / Partially applicable / Not applicable.
3
G.3.2
Blood pressure targets for treatment
7
0
B. Jonsson, L. Hansson, and N. O. Stalhammar. Health economics in the hypertension optimal treatment (HOT) study: costs and cost-effectiveness of intensive
blood pressure lowering and low-dose aspirin in patients with hypertension. J.Intern.Med. 253:472-480, 2003.
Study details
Economic
analysis: Cost
analysis/CCA
Study design:
Within RCT
analysis
Perspective:
International
resource use,
Population & interventions
Health outcomes
Costs
Cost effectiveness
Population:
Patients with hypertension and DBP110115mmHg
 N = 18,790
 Countries = 26
 Male = NR
 Mean age = 61.5years
None reported in paper for
whole population (only
diabetes subgroup). Paper
states that differences in
events were nonsignificant in the overall
population.
Total costs – all patients (mean
per patient):
Intvn 1: £2200
Intvn 2: £2282
Intvn3: £2381
Incremental (2-1):£82
(CI: NR , p<0.01)
Incremental (3-2):£99
(CI: NR , p<0.01)
Incremental (3-1):£181
Basecase ICER:
Costs significantly increased
as the target was lowered
and there was no significant
difference in cv events.
Subgroup analyses:
Diabetes subgroup – out of
scope of guideline.
Intervention 1: Treatment to target DBP
<90mmHg
See clinical evidence
review for more details.
Analysis of uncertainty:
Cost of non-cv
Hypertension (partial update)
Data sources
Evidence tables – health economic studies (2011 update)
incorporated:
Antihypertensive drugs,
physician visits, HBPM
Treatment intensified if DBP >89mmHg; treatment
reduced if DBP <80mmHg
(CI: NR , p<0.01)
Currency & cost year:
1995 Swedish Kroner (presented
here as 1995 UK pounds†)
All patients received felodipine (CCB). Additional
therapy and dose increments in four further steps
were prescribed to reach the randomised target.
50% of patients were also randomised to aspirin
(75mg daily).
hospitalisations included –
increased total costs,
differences remained
similar. Swedish patients
only used – not considered
relevant to guideline.
Cost components incorporated:
Antihypertensive drugs,
physician visits, hospitalisations,
side effects.
Data sources
Health outcomes: Within RCT analysis; Quality-of-life weights: N/a; Cost sources: Resource use collected within RCT; units costs were from Swedish national data
sources or published studies.
Comments
Source of funding: AstraZeneca; Limitations: Some uncertainty about applicability of International resource use and Swedish unit costs. QALYs not used (although
clinical outcomes reported as not significantly different). Discounting not applied. Within RCT analysis and so does not incorporate all available evidence on differences
between targets; issues raised with interpretation of clinical trial as achieved BPs very similar despite different targets. Other: n/a
Overall quality*: Potentially serious limitations
Overall applicability**: Partially applicable
3
7
1
Abbreviations: BP = blood pressure; CCA = cost-consequence analysis; CI = confidence interval; cv = cardiovascular; DBP = diastolic blood pressure; ICER = incremental costeffectiveness ratio; NR = not reported; RCT = randomised clinical trial; QALYs = quality-adjusted life years.
† Converted using 1995 Purchasing Power Parities468.
*Very serious limitations / Potentially serious Limitations / Minor limitations; ** Directly applicable / Partially applicable / Not applicable.
G.4
Pharmacological interventions
G.4.1
People aged over 80 years
T. D. Szucs, B. Waeber, and Y. Tomonaga. Cost-effectiveness of antihypertensive treatment in patients 80 years of age or older in Switzerland: an analysis of
the HYVET study from a Swiss perspective. Journal of Human Hypertension 24 (2):117-123, 2010.
Study details
Population & interventions
Health outcomes
Costs
Cost effectiveness
Economic analysis: CEA
Population:
Patients with hypertension
and aged >80 years.
Primary outcome
measure:
Life-years (mean per
patient)
Total costs – all patients (mean
per patient):
Intvn 1: £1021 (undiscounted)
Intvn 2: £1006 (undiscounted)
Basecase ICER:
Treatment dominated no treated
(lower costs and improved health
outcomes)
Study design:
Decision analytic model based on
Hypertension (partial update)
Intervention 2: Treatment to target DBP
<85mmHg
Intervention 3: Treatment to target DBP
<80mmHg
Evidence tables – health economic studies (2011 update)
Swedish costs,
healthcare payer
Followup: mean
3.8 years
Discounting:
Costs: NR;
Outcomes: NR
100% compliance assumed
for all patients.
Intvn 1: 1.6216
Intvn 2: 1.6672
Incremental (1-2):
0.0457
(CI: NR )
Other outcome
measures (mean per
patient):
None
Incremental (2-1):-£14 (discounted)
(CI: NR , p=NR)
Other: None
Subgroup analyses: None
Currency & cost year:
2007 Swiss Francs (presented here
as 2007 UK pounds†)
Analysis of uncertainty:
One way sensitivity analyses of
20% variation in medication cost,
cost of stroke, cost of HF, cost of
MI, life expectancy.
Cost components incorporated:
Antihypertensive drugs, acute
management and follow-up of MI,
stroke and heart failure
(medication, interventions,
hospitalisation, outpatient
treatment, rehabilitation).
Medication cost and cost of stroke
had the biggest impact.
Results varied from dominant to
£1097 per life year gained.
Data sources
Health outcomes: Swiss population mortality data and HYVET RCT639; Quality-of-life weights: N/a; Cost sources: Resource use for drugs based on HYVET with
units costs from Swiss pharmacy retail prices. MI, stoke and HF costs based on published studies – each event includes acute costs and 2 years of follow-up.
Comments
3
7
2
Source of funding: None; Limitations: Some uncertainty about applicability of Swiss unit costs. QALYs not used. Discounting not in line with NICE reference case.
Based on single RCT and so does not incorporate all available clinical evidence for patients over 80. Some methodological issues about how health outcomes and costs
are calculated and attributed in model. Other: n/a
Overall quality*: Potentially serious limitations
Overall applicability**: Partially applicable
Abbreviations: BP = blood pressure; CEA = cost effectiveness analysis; CI = confidence interval; DBP = diastolic blood pressure; HF = heart failure; ICER = incremental cost-effectiveness
ratio; MI = myocardial infarction; NR = not reported; RCT = randomised clinical trial; QALYs = quality-adjusted life years.
† Converted using 2007 Purchasing Power Parities468.
*Very serious limitations / Potentially serious Limitations / Minor limitations; ** Directly applicable / Partially applicable / Not applicable.
Hypertension (partial update)
Perspective: Switzerland,
healthcare payer
Time horizon: 2 years
Discounting: Costs: 5%;
Outcomes: None
Intervention 1:
No antihypertensive
treatment
Intervention 2:
Antihypertensive treatment
 25.8% 1.5mg indapamide
SR alone
 23.9% 1.5mg indapamide
SR and 2mg perindopril
 49.5% 1.5mg indapamide
SR and 4mg perindopril
Evidence tables – health economic studies (2011 update)
single RCT (HYVET639).
Approach to modelling:
Estimated costs based on drug
doses and usage in HYVET and
key clinical events. Life-years
gained were calculated by
applying the estimated life
expectancy for the population to
deaths avoided with treatment.
Hypertension (partial update)
Forest plots
Appendix H: Forest plots
H.1
Head to head comparisons, see section 10.2
The final cut-off date for all searches was 19 December 2005.
The following abbreviations were used: ACEi = angiotensin-converting inhibitors; ARB =
angiotensin-II receptor antagonists; BB = beta-blockers; CCB = calcium-channel blockers; CI
= confidence interval; MI = myocardial infarction; RR = relative risk.
Figure 20:
ACE Inhibitors versus Calcium Channel Blockers - Mortality
Figure 21:
ACE Inhibitors versus Calcium Channel Blockers – Myocardial
Infarction
373
Hypertension (partial update)
Forest plots
Figure 22:
ACE Inhibitors versus Calcium Channel Blockers - Stroke
Figure 23:
ACE Inhibitors versus Calcium Channel Blockers – Heart Failure
Figure 24:
ACE Inhibitors versus Calcium Channel Blockers - Diabetes
374
Hypertension (partial update)
Forest plots
Figure 25:
ARBs versus Beta-Blockers - Mortality
Figure 26:
ARBs versus Beta-Blockers – Myocardial Infarction
Figure 27:
ARBs versus Beta-Blockers - Stroke
375
Hypertension (partial update)
Forest plots
Figure 28:
ARBs versus Beta Blockers – Heart Failure
Figure 29:
ARBs versus Beta-Blockers - Diabetes
Figure 30:
ARBs versus Calcium Channel Blockers - Mortality
376
Hypertension (partial update)
Forest plots
Figure 31:
ARBs versus Calcium Channel Blockers – Myocardial Infarction
Figure 32:
ARBs versus Calcium Channel Blockers - Stroke
Figure 33:
ARBs versus Calcium Channel Blockers – Heart Failure
377
Hypertension (partial update)
Forest plots
Figure 34:
ACE Inhibitors versus Thiazides - Mortality
Figure 35:
ACE Inhibitors versus Thiazides – Myocardial Infarction
Figure 36:
ACE Inhibitors versus Thiazides - Stroke
378
Hypertension (partial update)
Forest plots
Figure 37:
ACE Inhibitors versus Thiazides – Heart Failure
Figure 38:
Calcium Channel Blockers versus Beta-Blockers - Mortality
Figure 39:
Calcium Channel Blockers versus Beta-Blockers – Myocardial Infarction
379
Hypertension (partial update)
Forest plots
Figure 40:
Calcium Channel Blockers versus Beta-Blockers – Myocardial Infarction
Figure 41:
Calcium Channel Blockers versus Beta-Blockers - Stroke
Figure 42:
Calcium Channel Blockers versus Beta-Blockers – Heart Failure
380
Hypertension (partial update)
Forest plots
Figure 43:
Calcium Channel Blockers versus Beta-Blockers - Diabetes
Figure 44:
Calcium Channel Blockers versus Thiazides - Mortality
Figure 45:
Calcium Channel Blockers versus Thiazides – Myocardial Infarction
381
Hypertension (partial update)
Forest plots
Figure 46:
Calcium Channel Blockers versus Thiazides - Stroke
Figure 47:
Calcium Channel Blockers versus Thiazdies – Heart Failure
382
Hypertension (partial update)
Forest plots
Figure 48:
Calcium Channel Blockers versus Thiazides - Diabetes
Figure 49:
Antihypertensive drug therapy versus placebo - Mortality
Figure 50:
Anti-hypertensive drug therapy versus placebo – Myocardial Infarction
383
Hypertension (partial update)
Forest plots
Figure 51:
H.1.1
Antihypertensive drug therapy versus placebo - Stroke
First line therapy, ACEi versus ARBs, see section 0
Figure 52:
Mortality
Figure 53:
Myocardial infarction
Figure 54:
Stroke
384
Hypertension (partial update)
Forest plots
Figure 55:
Hospitalisation for angina
Figure 56:
Coronary revascularisation
Figure 57:
New onset diabetes
Figure 58:
Heart Failure
385
Hypertension (partial update)
Forest plots
Figure 59:
H.1.2
Study drug withdrawals
ACEi versus ARBs, see section 10.3.1
Figure 60:
Mortality
Figure 61:
Myocardial Infarction
Figure 62:
Stroke
386
Hypertension (partial update)
Forest plots
Figure 63:
Hospitalisation for Angina
Figure 64:
Coronary revascularisation
Figure 65:
New onset diabetes
Figure 66:
Heart failure
387
Hypertension (partial update)
Forest plots
Figure 67:
Study drug withdrawal
H.1.3
Diuretics
H.1.3.1
Indapamide versus placebo
Figure 68:
Study or Subgroup
HYVET
PATS
Total (95% CI)
Overall mortality
Indapamide
placebo
Events Total Events Total
196
146
1933
2841
4774
O-E Variance Weight
235 1912 -22.43
158 2824 -6.58
4736
106.87
75.88
58.5%
41.5%
0.81 [0.67, 0.98]
0.92 [0.73, 1.15]
100.0%
0.85 [0.74, 0.99]
Total events
342
393
Heterogeneity: Chi² = 0.67, df = 1 (P = 0.41); I² = 0%
Test for overall effect: Z = 2.15 (P = 0.03)
Figure 69:
Study or Subgroup
HYVET
PATS
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.5
0.7
1
1.5
2
Favours indapamide Favours placebo
CHD event
Indapamide
placebo
Events Total Events Total O-E Variance Weight
22 1933
57 1912 -15.5
15.87 58.2%
25 2841
21 2824 -1.92
11.41 41.8%
Total (95% CI)
4774
4736
Total events
47
78
Heterogeneity: Chi² = 4.34, df = 1 (P = 0.04); I² = 77%
Test for overall effect: Z = 3.34 (P = 0.0009)
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.38 [0.23, 0.62]
0.85 [0.47, 1.51]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.53 [0.36, 0.77]
0.01 0.1
1
10
100
Favours indapamide Favours placebo
388
Hypertension (partial update)
Forest plots
Figure 70:
Study or Subgroup
HYVET
PATS
Stroke
indapamide
placebo
Events Total Events Total
51
159
Total (95% CI)
O-E Variance Weight
1933
2841
69 1912 -9.48
217 2824 -29.6
4774
4736
29.33
91.76
24.2%
75.8%
0.72 [0.50, 1.04]
0.72 [0.59, 0.89]
100.0%
0.72 [0.61, 0.87]
Total events
210
286
Heterogeneity: Chi² = 0.00, df = 1 (P = 1.00); I² = 0%
Test for overall effect: Z = 3.55 (P = 0.0004)
Figure 71:
Study or Subgroup
HYVET
PATS
Study or Subgroup
PATS
Cardiovascular event
Indapamide
placebo
Events Total Events Total
O-E Variance Weight
9 1933
12 1912 -1.53
5.14
4.5%
194 2841
247 2824 -28.11
108.66 95.5%
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.74 [0.31, 1.76]
0.77 [0.64, 0.93]
0.77 [0.64, 0.93]
0.2
0.5
1
2
5
Favours indapamide Favours placebo
indapamide
placebo
Events Total Events Total
2841
O-E Variance Weight
2019 2824 90.32 1,035.32 100.0%
2841
2824
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1.09 [1.03, 1.16]
1.09 [1.03, 1.16]
Total events
2125
2019
Heterogeneity: Not applicable
Test for overall effect: Z = 2.81 (P = 0.005)
H.1.3.2
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Quality of life
2125
Total (95% CI)
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.2
0.5
1
2
5
Favours indapamide Favours placebo
Total (95% CI)
4774
4736
Total events
203
259
Heterogeneity: Chi² = 0.01, df = 1 (P = 0.93); I² = 0%
Test for overall effect: Z = 2.78 (P = 0.005)
Figure 72:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.5
0.7
1
1.5
2
Favours placebo Favours indapamide
Chlorthalidone versus placebo
Figure 73:
Study or Subgroup
SHEP
SHEP-P
VA-NHLBI
Overall mortality
Chlorthalidone
placebo
Events
Total Events Total
O-E Variance Weight
213
2365
242 2371 -14.96
113.29 92.5%
32
443
7 108 -0.66
6.15
5.0%
8
508
5 504 -1.45
3.08
2.5%
Total (95% CI)
3316
2983
Total events
253
254
Heterogeneity: Chi² = 0.35, df = 2 (P = 0.84); I² = 0%
Test for overall effect: Z = 1.54 (P = 0.12)
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.88 [0.73, 1.05]
0.90 [0.41, 1.98]
0.62 [0.20, 1.91]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.87 [0.73, 1.04]
0.01
0.1
1
10
100
Favours chlorthalidone Favours placebo
389
Hypertension (partial update)
Forest plots
Figure 74:
Study or Subgroup
SHEP
SHEP-P
VA-NHLBI
CHD events
Chlorthalidone
placebo
Events Total Events Total
O-E Variance Weight
140
2365
184 2371 -23.18
79.51 92.0%
8
443
2 108 -0.05
1.58 1.8%
16
508
8 504 3.69
5.33 6.2%
Total (95% CI)
3316
2983
Total events
164
194
Heterogeneity: Chi² = 4.90, df = 2 (P = 0.09); I² = 59%
Test for overall effect: Z = 2.10 (P = 0.04)
Figure 75:
Study or Subgroup
SHEP
SHEP-P
Study or Subgroup
SHEP
VA-NHLBI
0.80 [0.65, 0.98]
0.5 0.7 1 1.5 2
Favours chlorthalidone Favours placebo
Chlorthalidone
placebo
Events Total Events Total
O-E Variance Weight
103
2365
159 2371 -27.52
62.51 95.9%
11
443
6 108 -2.69
2.68 4.1%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.64 [0.50, 0.83]
0.37 [0.11, 1.21]
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.63 [0.49, 0.80]
0.01
0.1
1
10
100
Favours chlorthalidone Favours placebo
Cardiovascular events
Chlorthalidone
placebo
Events
Total Events Total
O-E Variance Weight
199
2365
289 2371 -42.97
117.85 99.6%
2
508
0 504 0.73
0.5
0.4%
Total (95% CI)
2873
2875
Total events
201
289
Heterogeneity: Chi² = 1.66, df = 1 (P = 0.20); I² = 40%
Test for overall effect: Z = 3.88 (P = 0.0001)
H.1.3.3
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Stroke
Total (95% CI)
2808
2479
Total events
114
165
Heterogeneity: Chi² = 0.82, df = 1 (P = 0.37); I² = 0%
Test for overall effect: Z = 3.74 (P = 0.0002)
Figure 76:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.75 [0.60, 0.93]
0.97 [0.20, 4.61]
2.00 [0.86, 4.67]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.69 [0.58, 0.83]
4.31 [0.27, 68.84]
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.70 [0.58, 0.84]
0.01
0.1
1
10
100
Favours chlorthalidone Favours placebo
Chlorthalidone versus calcium channel blockers
Figure 77:
Study or Subgroup
ALLHAT
SHELL
VHAS
Overall mortality
Chlorthalidone
CCB
Events
Total Events
2203
14836
1256
122
940
145
4
707
5
Total O-E Variance Weight
8790 33.4
808.12 92.2%
942 -4.5
66.25
7.6%
707 0.49
2.22
0.3%
Total (95% CI)
16483
10439
Total events
2329
1406
Heterogeneity: Chi² = 0.81, df = 2 (P = 0.67); I² = 0%
Test for overall effect: Z = 0.99 (P = 0.32)
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1.04 [0.97, 1.12]
0.93 [0.73, 1.19]
1.25 [0.33, 4.65]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1.03 [0.97, 1.10]
0.5
0.7
1
1.5
Favours chlorthalidone Favours CCB
390
2
Hypertension (partial update)
Forest plots
Figure 78:
Study or Subgroup
ALLHAT
VHAS
CHD events
Chlorthalidone
CCB
Events
Total Events Total
2451
9
Total (95% CI)
O-E Variance Weight
14836
707
1466 8790 -62.2
8 707 4.23
15543
9497
917.33
4.24
99.5%
0.5%
0.93 [0.88, 1.00]
2.71 [1.05, 7.03]
100.0%
0.94 [0.88, 1.00]
Total events
2460
1474
Heterogeneity: Chi² = 4.79, df = 1 (P = 0.03); I² = 79%
Test for overall effect: Z = 1.91 (P = 0.06)
Figure 79:
Study or Subgroup
ALLHAT
SHELL
VHAS
0.5
0.7
1
1.5
Favours chlorthalidone Favours CCB
Chlorthalidone
CCB
Events
Total Events Total
675
38
4
14836
940
707
377
37
5
16483
O-E Variance Weight
8790 -16.8
942 0.54
707 -0.49
10439
241.9
18.75
2.22
ALLHAT
Total (95% CI)
Study or Subgroup
SHELL
0.93 [0.82, 1.06]
1.03 [0.65, 1.62]
0.80 [0.22, 2.99]
100.0%
0.94 [0.83, 1.06]
Study or Subgroup
SHELL
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.5
0.7
1
1.5
Favours chlorthalidone Favours CCB
Chlorthalidone
CCB
Events
Total Events Total
3941
14836
O-E Variance Weight
2432 8790 -60.29
14836
2
8790
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1,503.92 100.0%
0.96 [0.91, 1.01]
100.0%
0.96 [0.91, 1.01]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
2432
0.5
0.7
1
1.5
Favours chlorthalidone Favours CCB
2
Heart failure (fatal and non-fatal)
Chlorthalidone
CCB
Events
Total Events Total O-E Variance Weight
19
940
23
942 -1.98
10.4 100.0%
Total (95% CI)
940
Total events
19
Heterogeneity: Not applicable
Test for overall effect: Z = 0.61 (P = 0.54)
Figure 82:
2
Cardiovascular events
Total events
3941
Heterogeneity: Not applicable
Test for overall effect: Z = 1.55 (P = 0.12)
Figure 81:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
92.0%
7.1%
0.8%
Total events
717
419
Heterogeneity: Chi² = 0.22, df = 2 (P = 0.89); I² = 0%
Test for overall effect: Z = 1.03 (P = 0.30)
Study or Subgroup
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Stroke
Total (95% CI)
Figure 80:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
942
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.83 [0.45, 1.52]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.83 [0.45, 1.52]
23
0.01
0.1
1
10
100
Favours Chlorthalidone Favours CCB
Myocardial infarction
Chlorthalidone
CCB
Events
Total Events Total O-E Variance Weight
14
940
12
942 1.01
6.46 100.0%
Total (95% CI)
940
Total events
14
Heterogeneity: Not applicable
Test for overall effect: Z = 0.40 (P = 0.69)
942
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1.17 [0.54, 2.53]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1.17 [0.54, 2.53]
12
0.01
0.1
1
10
100
Favours chlorthalidone Favours CCB
391
Hypertension (partial update)
Forest plots
H.1.3.4
Chlorthalidone versus ACE Inhibitors
Figure 83:
Study or Subgroup
ALLHAT
ANBP2
Overall mortality
Chlorthalidone
ACEi
Events
Total Events Total
2203
210
Total (95% CI)
14836
3037
1314
195
O-E Variance Weight
8778 -3.61
3044 7.94
17873
823.07
101.11
11822
89.1%
10.9%
1.00 [0.93, 1.07]
1.08 [0.89, 1.31]
100.0%
1.00 [0.94, 1.07]
Total events
2413
1509
Heterogeneity: Chi² = 0.62, df = 1 (P = 0.43); I² = 0%
Test for overall effect: Z = 0.14 (P = 0.89)
Figure 84:
Study or Subgroup
ALLHAT
ANBP2
0.5
0.7
1
1.5
Favours chlorthalidone Favours ACE
Chlorthalidone
ACEi
Events
Total Events Total
2451
82
14836
3037
1505
58
17873
O-E Variance Weight
8778 -40.94
3044 11.97
932.45
33.97
11822
ALLHAT
ANBP2
0.96 [0.90, 1.02]
1.42 [1.02, 1.99]
100.0%
0.97 [0.91, 1.03]
0.2
0.5
1
2
5
Favours chlorthalidone Favours ACE
Chlorthalidone
ACEi
Events
Total Events Total
675
107
14836
3037
457
112
O-E Variance Weight
8778 -37.83
3044 -2.45
17873
264.37
54.72
11822
ALLHAT
ANBP2
Total (95% CI)
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
82.9%
17.1%
0.87 [0.77, 0.98]
0.96 [0.73, 1.25]
100.0%
0.88 [0.79, 0.98]
Total events
782
569
Heterogeneity: Chi² = 0.44, df = 1 (P = 0.51); I² = 0%
Test for overall effect: Z = 2.25 (P = 0.02)
Study or Subgroup
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Stroke
Total (95% CI)
Figure 86:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
96.5%
3.5%
Total events
2533
1563
Heterogeneity: Chi² = 5.15, df = 1 (P = 0.02); I² = 81%
Test for overall effect: Z = 0.93 (P = 0.35)
Study or Subgroup
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
CHD events
Total (95% CI)
Figure 85:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.5 0.7 1
1.5 2
Favours chlorthalidone Favours ACEi
Cardiovascular events
Chlorthalidone
ACEi
Events
Total Events Total
3941
429
14836
3037
2514
394
17873
O-E Variance Weight
8778 -124.24
3044 -19.21
11822
1,274
205.38
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
86.1%
13.9%
0.91 [0.86, 0.96]
0.91 [0.79, 1.04]
100.0%
0.91 [0.86, 0.96]
Total events
4370
2908
Heterogeneity: Chi² = 0.00, df = 1 (P = 0.96); I² = 0%
Test for overall effect: Z = 3.73 (P = 0.0002)
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.5
0.7
1
1.5
2
Favours Chlorthalidone Favours ACEi
392
2
Hypertension (partial update)
Forest plots
H.1.3.5
Hydrochlorthiazide versus calcium channel blockers
Figure 87:
Study or Subgroup
MIDAS
Sareli 2001
THAI elderly
Overall mortality
HCTZ
CCB
Events Total Events Total
9
1
0
Total (95% CI)
441
58
100
8
1
1
599
O-E Variance Weight
442 0.52
291 0.61
100 -0.34
4.24
0.28
0.25
833
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
88.9%
5.9%
5.2%
1.13 [0.44, 2.93]
8.83 [0.22, 358.72]
0.26 [0.01, 12.93]
100.0%
1.18 [0.48, 2.90]
Total events
10
10
Heterogeneity: Chi² = 1.72, df = 2 (P = 0.42); I² = 0%
Test for overall effect: Z = 0.36 (P = 0.72)
Figure 88:
Study or Subgroup
MIDAS
Sareli 2001
0.01 0.1
1
10
100
Favours HCTZ Favours CCB
CHD events
HCTZ
CCB
Events Total Events Total
12
1
Total (95% CI)
441
58
17
2
499
O-E Variance Weight
442 -2.48
291 0.49
733
7.03
0.42
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
94.4%
5.6%
0.70 [0.34, 1.47]
3.21 [0.16, 66.09]
100.0%
0.77 [0.37, 1.57]
Total events
13
19
Heterogeneity: Chi² = 0.91, df = 1 (P = 0.34); I² = 0%
Test for overall effect: Z = 0.73 (P = 0.47)
Figure 89:
Study or Subgroup
MIDAS
Study or Subgroup
MIDAS
Sareli 2001
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.01 0.1
1
10
100
Favours HCTZ Favours CCB
Stroke
CCB
HCTZ
Events Total Events Total O-E Variance Weight
6 442
3 441 -1.38
2 100.0%
Total (95% CI)
442
Total events
6
Heterogeneity: Not applicable
Test for overall effect: Z = 0.98 (P = 0.33)
Figure 90:
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
441
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.50 [0.13, 2.01]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.50 [0.13, 2.01]
3
0.01 0.1
1
10 100
Favours HCTZ Favours CCB
Cardiovascular events
HCTZ
CCB
Events Total Events Total O-E Variance Weight
14 441
25 442 -5.24
8.97 98.5%
0
58
1 291 -0.11
0.14
1.5%
Total (95% CI)
499
733
Total events
14
26
Heterogeneity: Chi² = 0.01, df = 1 (P = 0.94); I² = 0%
Test for overall effect: Z = 1.77 (P = 0.08)
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.56 [0.29, 1.07]
0.46 [0.00, 85.84]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.56 [0.29, 1.06]
0.01 0.1
1
10 100
Favours HCTZ Favours CCB
393
Hypertension (partial update)
Forest plots
H.1.3.6
Hydrochlorthiazide verus ACE Inhibitor
Figure 91:
Study or Subgroup
Sareli 2001
Overall mortality
HCTZ
ACEi
Events Total Events Total
1
Total (95% CI)
58
58
Total events
1
Heterogeneity: Not applicable
Test for overall effect: Z = 0.70 (P = 0.48)
Figure 92:
Study or Subgroup
PHYLLIS
3
253
Study or Subgroup
PHYLLIS
4.06 [0.08, 204.37]
100.0%
4.06 [0.08, 204.37]
0.001
0.1
1
10
1000
Favours HCTZ Favours ACE
1
253
O-E Variance Weight
254 0.83
254
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.75 100.0%
3.02 [0.31, 29.07]
100.0%
3.02 [0.31, 29.07]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
1
0.01 0.1
1
10
100
Favours HCTZ Favours ACE
Stroke
HCTZ
ACEi
Events Total Events Total O-E Variance Weight
0 253
1 254 -0.34
0.25 100.0%
Total (95% CI)
253
Total events
0
Heterogeneity: Not applicable
Test for overall effect: Z = 0.68 (P = 0.50)
Figure 94:
60
0.25 100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0
HCTZ
ACEi
Events Total Events Total
Total events
3
Heterogeneity: Not applicable
Test for overall effect: Z = 0.96 (P = 0.34)
Study or Subgroup
PHYLLIS
60 0.35
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
CHD events
Total (95% CI)
Figure 93:
0
O-E Variance Weight
254
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.26 [0.01, 12.93]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.26 [0.01, 12.93]
1
0.001
0.1 1 10
1000
Favours HCTZ Favours ACEi
Cardiovascular events
HCTZ
ACEi
Events Total Events Total O-E Variance Weight
0 253
1 254 -0.34
0.25 100.0%
Total (95% CI)
253
Total events
0
Heterogeneity: Not applicable
Test for overall effect: Z = 0.68 (P = 0.50)
254
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.26 [0.01, 12.93]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.26 [0.01, 12.93]
1
0.002 0.1 1 10
500
Favours HCTZ Favours ACEi
394
Hypertension (partial update)
Forest plots
H.1.3.7
Bendroflumethazide versus beta blockers
Figure 95:
Study or Subgroup
MRC
Overall mortality
Bendroflumethiazide
Beta blocker
Events
Total Events Total O-E Variance Weight
128
Total (95% CI)
3519
3519
Total events
128
Heterogeneity: Not applicable
Test for overall effect: Z = 0.61 (P = 0.54)
Figure 96:
Study or Subgroup
MRC
MRC
100.0%
1.08 [0.84, 1.39]
0.2
0.5
1
2
5
Favours bendroflumethiaz Favours beya blocker
119
3519
103
3519
3558 8.73
3558
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
55.21 100.0%
1.17 [0.90, 1.52]
100.0%
1.17 [0.90, 1.52]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
103
0.01
0.1
1
10
100
Favours bendrflumethiazid Favours beta blocker
Stroke
Bendroflumethiazide
Beta blocker
Events
Total Events Total
18
3519
42
3519
Total events
18
Heterogeneity: Not applicable
Test for overall effect: Z = 2.97 (P = 0.003)
Study or Subgroup
MRC
1.08 [0.84, 1.39]
CHD events
Total (95% CI)
Figure 98:
3558
61.94 100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
120
Total events
119
Heterogeneity: Not applicable
Test for overall effect: Z = 1.17 (P = 0.24)
Study or Subgroup
3558 4.82
Bendroflumethiazide Beta blocker
Events
Total Events Total O-E Variance Weight
Total (95% CI)
Figure 97:
120
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
O-E Variance Weight
3558 -10.53
3558
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
12.6 100.0%
0.43 [0.25, 0.75]
100.0%
0.43 [0.25, 0.75]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
42
0.01
0.1
1
10
100
Favours bendroflumethiaz Favours beta blockers
Cardiovascular events
Bendroflumethiazide Beta blocker
Events
Total Events Total O-E Variance Weight
140
3519
146 3558 -2.25
71.47 100.0%
Total (95% CI)
3519
Total events
140
Heterogeneity: Not applicable
Test for overall effect: Z = 0.27 (P = 0.79)
3558
100.0%
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.97 [0.77, 1.22]
Hazard Ratio
Exp[(O-E) / V], Fixed, 95% CI
0.97 [0.77, 1.22]
146
0.5
0.7
1
1.5
2
Favours Bendroflumethiaz Favours beta blocker
395
Hypertension (partial update)
Forest plots
H.1.4
Ethnicity, see section 10.7.2
Figure 99:
Angioedema
H.1.5
Diuretics
H.1.5.1
Indapamide versus cholrthalidone
Figure 100:
Study or Subgroup
Hatt 1975
Systolic blood pressure (end of follow-up)
Indapamide
Chlorthalidone
Mean Difference
Mean SD Total Mean SD Total Weight
IV, Fixed, 95% CI
153
21
Total (95% CI)
38
153
24
38
38 100.0% 0.00 [-10.14, 10.14]
38 100.0% 0.00 [-10.14, 10.14]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.00 (P = 1.00)
Figure 101:
Study or Subgroup
Hatt 1975
Total (95% CI)
Mean Difference
IV, Fixed, 95% CI
-20 -10 0 10 20
Favours indapamide Favours chlorthalidone
Diastolic blood pressure (end of follow-up)
Indapamide
Chlorthalidone
Mean Difference
Mean SD Total Mean SD Total Weight IV, Fixed, 95% CI
81
5
38
38
85
18
Mean Difference
IV, Fixed, 95% CI
38 100.0% -4.00 [-9.94, 1.94]
38 100.0% -4.00 [-9.94, 1.94]
Heterogeneity: Not applicable
Test for overall effect: Z = 1.32 (P = 0.19)
-20
-10
0
10
20
Favours indapamide Favours chlorthalidone
396
Hypertension (partial update)
Forest plots
Figure 102:
Study or Subgroup
Rakic 2002
Systolic blood pressure (change from baseline)
Indapamide
Chlorthalidone
Mean Difference
Mean
SD Total Mean
SD Total Weight IV, Fixed, 95% CI
-21.5 11.64
20
Total (95% CI)
-24.6
7.97
20
20 100.0%
3.10 [-3.08, 9.28]
20 100.0% 3.10 [-3.08, 9.28]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.98 (P = 0.33)
Figure 103:
Study or Subgroup
Rakic 2002
-100
-50
0
50
100
Favours indapamide Favours chlorthalidone
Diastolic blood pressure (change from baseline)
Indapamide
Chlorthalidone
Mean Difference
Mean SD Total Mean
SD Total Weight IV, Fixed, 95% CI
-10.1 5.82
Total (95% CI)
20
-13.6
6.17
20
20 100.0%
Mean Difference
IV, Fixed, 95% CI
3.50 [-0.22, 7.22]
20 100.0% 3.50 [-0.22, 7.22]
Heterogeneity: Not applicable
Test for overall effect: Z = 1.85 (P = 0.06)
H.1.5.2
Mean Difference
IV, Fixed, 95% CI
-100
-50
0
50
100
Favours indapamide Favours chlorthalidone
Indapamide versus hydrochlorthiazide
Figure 104:
Systolic blood pressure supine (end of follow-up)
Study or Subgroup
Indapamide
HCTZ
Mean
SD Total Mean SD Total Weight
Elliot 1991
Kreeft 1984
Plante 1983
Plante 1988
Spence 2000
153
9.8
142
11
127
4
162 14.37
136.1 14.9
Total (95% CI)
6
155 9.8
17
138
15
13
139
4
23
170 8.72
18 135.2 10.6
6
17
11
19
21
77
74 100.0%
Mean Difference
IV, Fixed, 95% CI
5.3%
-2.00 [-13.09, 9.09]
8.4%
4.00 [-4.84, 12.84]
63.5% -12.00 [-15.21, -8.79]
13.1% -8.00 [-15.06, -0.94]
9.6%
0.90 [-7.34, 9.14]
-8.36 [-10.92, -5.80]
Heterogeneity: Chi² = 18.56, df = 4 (P = 0.0010); I² = 78%
Test for overall effect: Z = 6.40 (P < 0.00001)
Figure 105:
Study or Subgroup
Elliot 1991
Kreeft 1984
Plante 1983
Plante 1988
Spence 2000
Total (95% CI)
Mean Difference
IV, Fixed, 95% CI
-20 -10
0
10 20
Favours indapamide Favours HCTZ
Diastolic blood pressure supine (end of follow-up)
Indapamide
HCTZ
Mean Difference
Mean SD Total Mean SD Total Weight IV, Fixed, 95% CI
93 4.9
88
7
83
2
89 9.58
87.1 5.32
6
17
13
23
18
94 4.9
86
7
89
2
94 8.72
88.3 3.9
77
6
17
11
19
21
Mean Difference
IV, Fixed, 95% CI
5.3% -1.00 [-6.54, 4.54]
7.4% 2.00 [-2.71, 6.71]
63.4% -6.00 [-7.61, -4.39]
5.3% -5.00 [-10.54, 0.54]
18.5% -1.20 [-4.17, 1.77]
74 100.0% -4.20 [-5.48, -2.92]
Heterogeneity: Chi² = 16.77, df = 4 (P = 0.002); I² = 76%
Test for overall effect: Z = 6.44 (P < 0.00001)
-20
-10
0
10
20
Favours indapamide Favours chlorthalidone
397
Hypertension (partial update)
Forest plots
Figure 106:
Systolic blood pressure upright (end of follow-up)
Study or Subgroup
Elliot 1991
Kreeft 1984
Plante 1983
Spence 2000
Indapamide
HCTZ
Mean Difference
Mean
SD Total Mean SD Total Weight
IV, Fixed, 95% CI
156 12.25
6 157 9.8
6
5.8% -1.00 [-13.55, 11.55]
137
16
17 137 17
17
7.4% 0.00 [-11.10, 11.10]
127
5
13 139 4
11 69.9% -12.00 [-15.60, -8.40]
136.3 15.8
18 138 1.4
21 16.9%
-1.70 [-9.02, 5.62]
Total (95% CI)
54
Heterogeneity: Chi² = 10.54, df = 3 (P = 0.01); I² = 72%
Test for overall effect: Z = 5.69 (P < 0.00001)
Figure 107:
55 100.0% -8.74 [-11.75, -5.73]
-20
-10
0
10
20
Favours indapamide Favours HCTZ
Diastolic blood pressure upright (end of follow-up)
Study or Subgroup
Elliot 1991
Kreeft 1984
Plante 1983
Spence 2000
Indapamide
HCTZ
Mean Difference
Mean SD Total Mean SD Total Weight
IV, Fixed, 95% CI
93
90
84
98.6
Total (95% CI)
4.9
7
2
5.9
6
17
13
18
94 4.9
91
7
90
3
99.7 3.7
54
6
17
11
21
55 100.0% -3.85 [-5.41, -2.28]
-20
-10
0
10
20
Favours indapamide Favours HCTZ
Systolic blood pressure supine (change from baseline)
Indapamide
HCTZ
Mean Difference
Mean SD Total Mean SD Total Weight
IV, Fixed, 95% CI
-22.7 15.4
178 -19.4 15.5
171 90.4%
-3.30 [-6.54, -0.06]
-19.1 18.7
18 -9.05 11.5
21
9.6% -10.05 [-19.99, -0.11]
Study or Subgroup
Emeriau 2001
Spence 2000
Total (95% CI)
196
Heterogeneity: Chi² = 1.60, df = 1 (P = 0.21); I² = 38%
Test for overall effect: Z = 2.51 (P = 0.01)
Figure 109:
Mean Difference
IV, Fixed, 95% CI
7.9% -1.00 [-6.54, 4.54]
11.0% -1.00 [-5.71, 3.71]
56.5% -6.00 [-8.08, -3.92]
24.6% -1.10 [-4.25, 2.05]
Heterogeneity: Chi² = 9.45, df = 3 (P = 0.02); I² = 68%
Test for overall effect: Z = 4.83 (P < 0.00001)
Figure 108:
Mean Difference
IV, Fixed, 95% CI
192 100.0%
Mean Difference
IV, Fixed, 95% CI
-3.95 [-7.03, -0.87]
-20 -10 0 10 20
Favours Indapamide Favours HCTZ
Diastolic blood pressure supine (change from baseline)
Study or Subgroup
Emeriau 2001
Spence 2000
Indapamide
HCTZ
Mean Difference
Mean SD Total Mean SD Total Weight IV, Fixed, 95% CI
-11.8 9.7 178 -10.8 8.5 171 82.8% -1.00 [-2.91, 0.91]
-9.4 7.7
18
-9.8 5.2
21 17.2% 0.40 [-3.80, 4.60]
Total (95% CI)
196
Heterogeneity: Chi² = 0.35, df = 1 (P = 0.55); I² = 0%
Test for overall effect: Z = 0.86 (P = 0.39)
Mean Difference
IV, Fixed, 95% CI
192 100.0% -0.76 [-2.50, 0.98]
-20
-10
0
10
20
Favours indapamide Favours HCTZ
398
Hypertension (partial update)
Forest plots
Figure 110:
Systolic blood pressure upright (change from baseline)
Study or Subgroup
Indapamide
HCTZ
Mean SD Total Mean SD Total Weight
Spence 2000
-18.5
1.1
18
Total (95% CI)
-5.95 10.6
18
Mean Difference
IV, Fixed, 95% CI
Mean Difference
IV, Fixed, 95% CI
21 100.0% -12.55 [-17.11, -7.99]
21 100.0% -12.55 [-17.11, -7.99]
Heterogeneity: Not applicable
Test for overall effect: Z = 5.39 (P < 0.00001)
Figure 111:
-20 -10 0 10 20
Favours HCTZ Favours indapamide
Diastolic upright (change from baseline)
Study or Subgroup
Spence 2000
Indapamide
HCTZ
Mean Difference
Mean SD Total Mean SD Total Weight IV, Fixed, 95% CI
-7.23 9.4
18 -5.16 6.4
21 100.0% -2.07 [-7.20, 3.06]
Total (95% CI)
18
Heterogeneity: Not applicable
Test for overall effect: Z = 0.79 (P = 0.43)
H.1.5.3
Mean Difference
IV, Fixed, 95% CI
21 100.0% -2.07 [-7.20, 3.06]
-20
-10
0
10
20
Favours HCTZ Favours indapamide
Indapamide versus Bendrofluazide/Bendroflumethiazide
Figure 112:
Study or Subgroup
Bing 1981
Systolic blood pressure supine (end of follow-up)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight
IV, Fixed, 95% CI
130 56.9
10
Total (95% CI)
162
31.6
10
10 100.0% -32.00 [-72.34, 8.34]
10 100.0% -32.00 [-72.34, 8.34]
Heterogeneity: Not applicable
Test for overall effect: Z = 1.55 (P = 0.12)
Figure 113:
Study or Subgroup
Bing 1981
-20 -10 0
10 20
Favours indapamide Favours bendrofluazide/be
Systolic blood pressure upright (end of follow-up)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight
IV, Fixed, 95% CI
148 37.9
Total (95% CI)
10
150
31.6
10
Study or Subgroup
Bing 1981
Total (95% CI)
Mean Difference
IV, Fixed, 95% CI
10 100.0% -2.00 [-32.58, 28.58]
10 100.0% -2.00 [-32.58, 28.58]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.13 (P = 0.90)
Figure 114:
Mean Difference
IV, Fixed, 95% CI
-20 -10
0
10
20
Favours indapamide Favours bendrofluazide/be
Diastolic blood pressure supine (end of follow-up)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight
IV, Fixed, 95% CI
85 15.8
10
10
90
15.8
Mean Difference
IV, Fixed, 95% CI
10 100.0% -5.00 [-18.85, 8.85]
10 100.0% -5.00 [-18.85, 8.85]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.71 (P = 0.48)
-20 -10 0
10 20
Favours indapamide Favours bendrofluazide/be
399
Hypertension (partial update)
Forest plots
Figure 115:
Study or Subgroup
Bing 1981
Diastolic blood pressure upright (end of follow-up)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight
IV, Fixed, 95% CI
100 47.4
Total (95% CI)
10
100
15.8
10
10 100.0% 0.00 [-30.97, 30.97]
10 100.0% 0.00 [-30.97, 30.97]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.00 (P = 1.00)
Figure 116:
Study or Subgroup
Alem 2008
-20 -10 0 10 20
Favours indapamide Favours bendrofluazide/be
Systolic blood pressure (absolute change)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight IV, Fixed, 95% CI
18.8 14.1
Total (95% CI)
13
13.2
18.8
13
Study or Subgroup
Alem 2008
Total (95% CI)
Mean Difference
IV, Fixed, 95% CI
10 100.0% 5.60 [-8.35, 19.55]
10 100.0% 5.60 [-8.35, 19.55]
Heterogeneity: Not applicable
Test for overall effect: Z = 0.79 (P = 0.43)
Figure 117:
Mean Difference
IV, Fixed, 95% CI
-20 -10
0
10
20
Favours bendrofluazide/be Favours indapamide
Diastolic blood pressure (absolute change)
Indapamide
Bendrofluazide
Mean Difference
Mean SD Total Mean
SD Total Weight IV, Fixed, 95% CI
8.6
6.4
13
13
5.4
5.9
10 100.0%
Mean Difference
IV, Fixed, 95% CI
3.20 [-1.85, 8.25]
10 100.0% 3.20 [-1.85, 8.25]
Heterogeneity: Not applicable
Test for overall effect: Z = 1.24 (P = 0.21)
-20
-10
0
10
20
Favours bendrofluazide/be Favours indapamide
400
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Appendix I: Cost-effectiveness analysis –
pharmacological treatment (updated 2011)
Introduction
This model was developed as part of the 2006 pharmacological update (CG34) to balance
clinical outcomes and to test the cost effectiveness of different initial antihypertensive
medications. As part of the 2011 update this analysis was rerun with updated costs. The
relative risks for ARBs were also updated based on new ACEi vs ARB data. The methods and
results below have been updated to reflect the revisions to the analysis. A summary of the
overall impact of the update compared to the previous analysis is given after the results
section.
I.2
Economic question
The aim of the model was to estimate the cost effectiveness of the various blood pressurelowering drug classes for the management of hypertension in primary care.
I.3
Methods
I.3.1
Population and subgroups
The model considered patients with essential hypertension seen in primary care, excluding
those with pre-existing cardiovascular disease (CVD), heart failure or diabetes. It was
designed to be run separately for different cohorts, defined by age (55, 65, 75 and 85) and sex.
In addition, the model classified these cohorts by baseline CVD risk (0.5–5% per year), by
heart failure risk (0–5% per year) and by diabetes risk (0–5% per year).
The basecase analysis presented below shows the results for 65-year-old men and women
with 2% CVD risk, 1% heart failure risk and 1.1% diabetes risk. Sensitivity analysis are also
presented showing whether and how the results vary by age, sex, CVD, heart failure and
diabetes risk.
The model is based on trial evidence that included relatively few younger (under 55) or black
people of African and Caribbean descent, so the results may not be reliable for these groups.
However, speculative sensitivity analyses were conducted to explore how different
assumptions about treatment effects might impact on the cost-effectiveness results for
younger people (under 45).
I.3.2
Interventions compared
The analysis assessed the costs and effects of alternative drugs alongside a 'do nothing'
comparator. Inclusion of no treatment as an option is important for economic evaluations as it
allows identification of low-risk groups for whom treatment is not likely to be cost effective.
The interventions compared were thus:
 no intervention (NI)
 thiazide-type diuretics (D)
 calcium-channel blockers (C)
 beta-blockers (B)
 angiotensin-converting enzyme inhibitors (ACEis) / angiotensin-II receptor blockers
(ARBs) (A).
401
Update 2011
I.1
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
At basecase, it was assumed that 80% of patients will be on ACEis and 20% will be on ARBs,
because of some people’s inability to tolerate ACEis (expert opinion). ACEi/ARBs were
combined as a strategy as they were considered to have equivalent effectiveness. The costs
and effects of the drugs were weighted to take account of this.
For simplicity, only first-line drugs were considered. However, it should be noted that the
relative treatment effects from the meta-analysis include additional benefits from various
second- and third-line treatments offered in the trials.
I.3.3
Outcomes
The treatment effects were measured in terms of prevention of CVD events: non-fatal
unstable angina, myocardial infarction (MI), heart failure and stroke, and CVD-related deaths.
Other CVD events, including onset of stable angina, peripheral vascular disease and transient
ischaemic attacks were not modelled, because data on them are not consistently reported in
the trials.
The only side effects modelled were onset of heart failure and diabetes. Other side effects
were not modelled in the basecase analysis, although the possible impact of adverse reactions
to the drugs in sensitivity analyses was examined. It should also be noted that the model does
not explicitly include cost impacts of withdrawals, non-concordance or transfers between
treatments. The impact of such changes on effectiveness is implicitly included through the use
of intention-to-treat trial data.
Health outcomes for the cost-effectiveness analysis are summarised in the form of qualityadjusted life-years (QALYs), where one QALY represents one year of healthy life.
I.3.4
Model structure and assumptions
A Markov model was developed to evaluate the incremental costs and effects of lifetime
treatment with alternative drugs for the management of hypertension in primary care from a
UK NHS perspective.
In a Markov model there are a finite number of health states. It is assumed that at any point in
time, all patients must be in one and only one of the states. The model then replicates how a
hypothetical cohort of people move between the states. Figure 118 shows a schematic
representation of the patients' pathways. All patients start in the event-free health state.
During each six-month cycle of the model, a proportion of patients enter one of the qualifying
event health states (MI, unstable angina, stroke, diabetes, heart failure or death) while the
remainder stay in the event-free state. Patients can experience more than one non-fatal event
in subsequent periods of the model. Ultimately, all patients end up in the death state.
402
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Figure 118:
a)
Model structure for hypertension(a)
Arrows represent the possible transitions between each of the health states
The rate at which people move through the model is regulated by transition probabilities,
which describe the likelihood of moving between states over each model cycle (6 months).
These transition probabilities are adjusted for each subgroup by age, sex, ethnicity, baseline
CVD, heart failure risk and diabetes risk. For illustration, the equivalent annual transition
probabilities for untreated 65-year-old men and women with 2% CVD, 1% heart failure risk
and 1.1% diabetes risk are shown in Table 82 and Table 83. Unless better data for a
hypertensive population were available, the probabilities were based on those used in an
analysis of the cost effectiveness of statins developed by the University of Sheffield's School
of Health and Related Research (ScHARR) for the NICE technology appraisal625. The GDG
advised on this and other data used in the model.
Table 82: Probabilities for a 65-year-old untreated man with 2% annual CVD risk
Parameter
Annual probability
(%)
Source
Well to unstable angina
0.0017
Statins model625
Well to MI
0.0035
Statins model625
Well to diabetes
0.0110
ASCOT trial157
Well to stroke
0.0054
Statins model625
Well to heart failure
0.0098
SHEP483
Well to death
0.0180
Statins model and population life tables625
Unstable angina to MI
0.0300
Statins model625
Unstable angina to diabetes
0.0067
Assumed to be the same as MI to diabetes
Unstable angina to stroke
0.0095
Assumed to be the same as MI to stroke
Unstable angina to heart failure
0.0230
Assumed to be the same as MI to heart failure
Unstable angina to death
0.0348
Statins model and population life tables625
MI to unstable angina
0.0078
HOPE25
MI to MI
0.0721
Statins model625
MI to diabetes
0.0067
HOPE25
MI to stroke
0.0095
Statins model625
MI to heart failure
0.0230
HOPE25
MI to death
0.0258
Statins model and population life tables625
Diabetes to unstable angina
0.0033
Double the risk of the well population
403
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Source
Diabetes to MI
Annual probability
(%)
0.0069
Diabetes to stroke
0.0108
Double the risk of the well population
Diabetes to heart failure
0.0197
Double the risk of the well population
Diabetes to death
0.0359
Double the risk of the well population
Stroke to unstable angina
0.0016
Assumed to be the same as stroke to MI
Stroke to MI
0.0016
Statins model625
Stroke to diabetes
0.0067
Assumed to be the same as MI to diabetes
Stroke to stroke
0.2875
Statins model625
Stroke to heart failure
0.0115
Assumed to be half of MI to heart failure
Stroke to death
0.3548
Statins model and population life tables625
Heart failure to unstable angina
0.0230
Assumed to be the same as heart failure to MI
Heart failure to MI
0.0230
SOLVD12
Heart failure to stroke
0.0103
SOLVD12
heart failure to heart failure
0.0545
SOLVD12
Heart failure to death
0.0768
SOLVD and population life tables12
Parameter
Double the risk of the well population
Table 83: Probabilities for a 65-year-old untreated woman with 2% annual CVD risk
Source
Well to unstable angina
Annual probability
(%)
0.0010
Well to MI
0.0024
Statins model625
Well to diabetes
0.0110
ASCOT trial157
Well to stroke
0.0076
Statins model625
Well to heart failure
0.0098
SHEP483
Well to death
0.0141
Statins model and population life tables625
Unstable angina to diabetes
0.0067
Assumed to be the same as MI to diabetes
Unstable angina to stroke
0.0095
Assumed to be the same as MI to stroke
Unstable angina to heart failure
0.0230
Assumed to be the same as MI to heart failure
Unstable angina to death
0.0307
Statins model and population life tables625
MI to unstable angina
0.0078
HOPE25
MI to MI
0.0721
Statins model625
MI to diabetes
0.0067
HOPE25
MI to stroke
0.0095
Statins model625
MI to HF
0.0230
HOPE25
MI to death
0.0217
Statins model and population life tables625
Diabetes to unstable angina
0.0021
Double the risk of the well population
Diabetes to MI
0.0048
Double the risk of the well population
Diabetes to stroke
0.0153
Double the risk of the well population
Diabetes to heart failure
0.0196
Double risk of well
Diabetes to death
0.0283
Double the risk of the well population
Stroke to unstable angina
0.0016
Assumed to be the same as stroke to MI
Stroke to MI
0.0016
Statins model625
Stroke to diabetes
0.0067
Assumed to be the same as MI to diabetes
Stroke to stroke
0.2875
Statins model625
Stroke to heart failure
0.0115
Assumed to be half of heart failure to MI
Parameter
404
Statins model625
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Source
Stroke to death
Annual probability
(%)
0.3507
Heart failure to unstable angina
0.023
Same as MI to heart failure
Heart failure to MI
0.023
SOLVD12
Heart failure to stroke
0.0103
SOLVD12
Heart failure to heart failure
0.0545
SOLVD12
Heart failure to death
0.0727
SOLVD and population life tables12
Parameter
Statins model and population life tables625
The model is run first assuming that the cohort was to receive no intervention (NI). The
model is then re-run for each active treatment (A, B, C and D) with transition probabilities
adjusted to reflect the expected reduction in CVD events and diabetes and heart failure
incidence from the clinical meta-analysis. Healthcare costs and QALYs are then estimated for
each option (NI, A, B, C and D) by weighting the time spent in the various states by mean
costs and 'utilities' (health-related quality of life) of the health states. The cost and utility data
used in the model are described below.
The time horizon modelled is a lifetime, with an assumed upper age of 100, by which time
most of the cohort have died.
I.3.5
Baseline risks
The probabilities of primary CVD events by age for a 45-year-old cohort with initial CVD
risk of 2% are shown in Table 84. CVD risk was assumed to rise at the rate of 0.03% per
annum for men and 0.008% per annum for women (estimated from the Health Survey for
England data 1998 by ScHARR24). The proportion of first CVD events that were unstable
angina, MI, stroke or death were taken from the age-specific UK incidence rates used in the
ScHARR statins model. In the statins model they obtained their data from the Bromley
Coronary Heart Disease Register and Oxfordshire Community Stroke Project. The risk of
new-onset diabetes in the baseline model (1.1%) was taken from the metabolically neutral
arm of the ASCOT trial157. The incidence of heart failure in the baseline model (0.98%) was
taken from the placebo arm of the SHEP trial483.
Table 84: Baseline incidences of primary events in untreated population
Distribution of primary cardiovascular disease events
Men
Age
UA %
MI %
Stroke %
CVD death %
Other* %
45
10.7
29.5
12.9
10.1
36.8
55
7.1
17.2
20.6
13.4
41.7
65
8.3
17.3
27.0
16.0
31.4
75
8.1
16.1
34.3
14.3
27.2
85
9.6
18.6
35.1
13.7
23.0
Source: ScHARR statins model625.
Women
Age
UA %
MI %
Stroke %
CVD death %
Other* %
45
11.7
8.0
22.9
9.1
48.3
55
7.3
9.2
28.8
10.6
44.1
65
5.2
12.1
38.2
17.1
27.4
75
3.4
10.2
46.4
15.2
24.8
85
2.9
10.0
50.1
14.7
22.3
Annual probability of primary cardiovascular disease events
Men
405
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Age
UA %
MI %
Stroke %
CVD death %
Total risk %
45
0.21
0.59
0.26
0.20
2.00
55
0.16
0.40
0.47
0.31
2.30
65
0.22
0.45
0.70
0.42
2.60
75
0.23
0.47
0.99
0.41
2.90
85
0.31
0.60
1.12
0.44
3.20
Age
UA %
MI %
Stroke %
CVD death %
Total risk %
45
0.23
0.16
0.46
0.18
2.00
55
0.15
0.19
0.60
0.22
2.08
65
0.11
0.26
0.83
0.37
2.16
75
0.08
0.23
1.04
0.34
2.24
85
0.07
0.23
1.16
0.34
2.32
Women
* Stable angina and TIA. UA = unstable angina; MI = myocardial infarction
The risk of CVD-related mortality was estimated from CVD incidence in the cohort, and the proportion of CVD events
estimated to be fatal (from the ScHARR model). Non-CVD related mortality by age and sex (Table 85) was taken from life
tables for England and Wales prepared by the Government Actuaries Department (GAD) and from data on the proportion of
deaths due to CVD-related causes from the Office for National Statistics457. In the base-case model it was assumed that the
hypertensive cohort was not at increased risk of non-CVD death compared with the general population. However, this
assumption was tested in a sensitivity analysis LC = lowest cost option; ’-‘ = option ruled out by simple or extended
dominance
* 1 = A and B dominated: C versus D
** Relative risk of CVD event following stroke compared with CVD event risks for people who have not had a CVD event
Risk of non-CVD death
As shown in Table 105, conclusions are not sensitive to changes in the assumptions about the
relative risk of death from non-CVD in the hypertensive cohort compared with the general
population. Hypertensive treatment remains highly cost-effective, and CCBs remain the
preferred option (holding all other variables at their base-case values).
Table 105, raising the cohort's relative risk from 1 to 8.
Table 85: Baseline non-cardiovascular disease related death
Deaths by age, sex and underlying cause, 2004 registrations, England and Wales
Ag
e
All cause ICD10: A00–R99
Circulatory ICD: I00–I99
Men
Women
Men
Women
Non-circulatory as proportion
of all deaths
Men %
Women %
45
12,417
8,139
3,930
1,362
0.68
0.83
55
27,117
17,649
9,330
3,541
0.66
0.80
65
52,709
37,041
19,783
11,304
0.62
0.69
75
87,367
88,404
35,607
35,958
0.59
0.59
85
51,329
109,488
20,816
46,470
0.59
0.58
Source: Office for National Statistics
457
All cause mortality, estimated from life tables, 2002–04, England and Wales
Age
Annual probability of death in age band
Men %
Women %
45
0.0037
0.0025
55
0.0093
0.0059
65
0.0236
0.0154
75
0.0537
0.0406
85
0.0870
0.0807
406
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Source: Government Actuary's Department458
Estimated non-circulatory deaths for hypertensive cohort
Age
Annual probability of death in age band
Men %
Women %
45
0.25%
0.20%
55
0.61%
0.47%
65
1.48%
1.07%
75
3.18%
2.41%
85
5.17%
4.65%
The risk of secondary or subsequent events, following unstable angina, MI, stroke or heart
failure are shown in Table 86. The increased risks of mortality and other CVD events for
patients who develop diabetes were assumed to be twice those seen in non-diabetic patients.
The British Hypertension Society guideline (2004)642 noted that the increase in CVD risk in
men is twice, while in women it is four-fold. This assumption was tested in a sensitivity
analysis. Probabilities of having unstable angina, heart failure and diabetes after an MI were
taken from HOPE25, which was a secondary prevention trial. The probability of having
diabetes after a stroke was assumed to be the same as that of having diabetes from MI. The
probabilities of unstable angina (UA), MI, stroke, heart failure and CVD death following
onset of heart failure were taken from the placebo arm of the SOLVD trial12. Because of a
lack of data, it was also assumed that transitions from UA to diabetes, heart failure and stroke
and from stroke to unstable angina were the same as those seen in the MI population (expert
opinion). It was also assumed that the risk of heart failure following a stroke is half that
following MI.
Table 86: Baseline incidences of secondary events in untreated population
After
Transition to
Annual risk
Source
Unstable angina (UA)
UA
No recurrence
Expert opinion
MI
0.03000
Statins model625
Diabetes
0.00667
Assumed same as MI to diabetes
Stroke
0.00950
Assumed same as MI to stroke
Heart failure
0.02300
Assumed same as MI to heart failure
CVD death
0.02000
Statins model625
UA
0.00775
HOPE25
MI
0.07210
Statins model625
Diabetes
0.00667
HOPE25
Stroke
0.00950
Statins model625
Heart failure
0.02300
HOPE25
CVD death
0.01100
Statins model625
UA
0.00160
Assumed same as for stroke to MI
MI
0.00160
Statins model625
Diabetes
0.00667
Assumed same as MI to diabetes
Stroke
0.28750
Statins model625
Heart failure
0.01150
Assumed half rate for MI to heart failure
CVD death
0.34000
Statins model625
UA
0.02300
SOLVD12
MI
0.02300
SOLVD12
Stroke
0.01025
SOLVD12
Heart failure
0.05450
SOLVD12
MI
Stroke
Heart failure
407
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
After
Transition to
Annual risk
Source
CVD death
0.06200
SOLVD12
I.3.6
Treatment effects
The relative treatment effects of these interventions were taken from the meta-analysis of
head-to-head studies done for the 2006 guideline update (except for the ACEi versus ARB
data that was taken from a meta analysis of studies in the 2011 update552,587,653. Comparisons
including data from large recent studies were chosen to estimate the treatment effects for the
economic evaluation: D versus NI, C versus D, C versus B, C versus ACEi, and ACEi versus
ARB (Table 87). Sensitivity analyses were conducted for two other scenarios: firstly by
replacing the estimate for B with a comparison with D (Table 88) and secondly by replacing
the estimate for ACEis with a comparison with D (Table 89).
Table 87: Relative risks of drugs versus no treatment (basecase analysis)
Calcium-channel
blockers (C)
0.881
Beta-blockers (B)
ACEi/ARB (A)
UA
Thiazide-type
diuretics (D)
0.893
0.984
1.01
MI
0.780
0.796
0.855
0.85
Diabetes
0.985
0.808
1.137
0.77
Stroke
0.690
0.656
0.851
0.69
Heart failure
0.530
0.731
0.761
0.65
Death
0.910
0.883
0.939
0.90
Outcome
Table 88: Relative risks of drugs versus no treatment (scenario 1: B versus D)
Outcome
Thiazide-type
diuretics (D)
Calcium-channel
blockers (C)
Beta-blockers (B)
ACEi/ARB (A)
UA
0.893
0.881
0.984 *
1.01
MI
0.780
0.796
0.835
0.85
Diabetes
0.985
0.808
1.138 *
0.77
Stroke
0.690
0.656
0.794
0.69
Heart failure
0.530
0.731
0.762 *
0.65
Death
0.910
0.883
0.901
0.90
* Based on B versus C comparison, since B versus D were not available for this outcome.
Table 89: Relative risks of drugs versus no treatment (scenario 2: ACEi versus D)
Calcium-channel
blockers (C)
0.881
Beta-blockers (B)
ACEi/ARB (A)
UA
Thiazide-type
diuretics (D)
0.893
0.984
1.01
MI
0.780
0.796
0.855
0.90
Diabetes
0.985
0.808
1.138
0.85
Stroke
0.690
0.656
0.851
0.64
Heart failure
0.530
0.731
0.762
0.72
Death
0.910
0.883
0.939
0.84
Outcome
I.3.7
Cost data
The NICE reference case430 specifies that costs should be measured from an NHS and
personal social services perspective. These should include the direct cost of drug treatment
and also potential savings from avoided treatments due to reduced incidence of CVD and/or
metabolic disease. Costs were calculated using cost weights for each of the states of the
408
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
model, multiplied by the time spent in each state. As per current NICE guidance430, an annual
discount rate of 3.5% was used for both costs and health benefits.
The costs of health states used in the model are shown in Table 90. Event costs reviewed as
part of the diagnosis model undertaken for the 2011 update were updated on the same basis
(stroke, MI, unstable angina) adjusted for the 6-month cycle length of this model (see
‘Appendix J: Cost-effectiveness analysis – blood pressure monitoring for confirming a
diagnosis of hypertension (new 2011)’ for details). Other event costs were simply inflated to
2009/10 costs (diabetes, heart failure).
Costs for stroke and post-stroke were based on Youman et al.649 (these were also used in the
NICE Statins health technology assessment (HTA)625). Costs of diabetes were based on
estimates from a NICE submission done by ScHARR when they evaluated the use of
sibutramine for the treatment of obesity48. Acute MI costs are based on those reported by
Palmer et al. (also utilised in the Statins HTA) this included costs for revascularisation. PostMI costs are based on an estimate made by the 2004 GDG for the hypertension guideline.
Initital and subsequent costs for unstable angina were assumed to be 60% of the costs for MI.
Heart failure costs were taken from NHS reference costs. It was assumed that people with no
event had an annual GP check-up (as recommended in the guideline). A check-up was
estimated to cost £28 based on the average UK cost of a GP appointment486.
Costs were inflated to 2009/10 prices using the Personal Social Services Research Unit
(PSSRU), Unit Costs of Health and Social Care 2010 inflation indices486.
Table 90: Costs of health states
Cost per
6-month cycle
Source
MI
£4,933
Palmer 2004474 inflated to 2009/10486
£282
NICE Hypertension guideline 2004441
inflated to 2009/10486
Unstable angina
£2,960
Assumed 60% of MI cost.
Subsequent unstable angina costs
£169
Assumed 60% of post-MI cost.
Diabetes
£455
Ara 200448 inflated to 2009/10486
Stroke
£10,190
Youman et al.649 inflated to 2009/10486
Post-stroke costs
£1,119
Youman et al.649 inflated to 2009/10486
£2,649
NHS reference costs 2005/06 inflated to
2009/10486
Post-heart failure costs
£282
Assumed to be same as post MI
Death
£0
No event
£28
Post-MI costs
Heart failure
Drug costs were calculated based on the prices quoted in the British National Formulary 60
(September 2010) based on the optimal dose for hypertension306. Optimal doses were
provided by clinical members of the GDG. In the base-case analysis, the non-proprietary cost
for the most commonly used drug in each class (as based on the 2008 NHS Prescription Cost
Analysis593) was used. The exception was for ARBs where losartan was used in the base-case
analysis as it has recently come off patent and so is now considerably cheaper than other
drugs in the class and therefore likely to be more commonly prescribed in the future than
historically. The impact of using the cheapest and most expensive drug in each class was also
tested in sensitivity analyses. Drug costs used are summarised in Table 91.
Table 91: Drug costs per year
ACEi
Used in basecase analysis
Cheapest drug
Most expensive drug
Ramipril (10mg): £20.73
Ramipril (10mg): £20.71
Cilaipril (5mg): £163.08
409
Update 2011
Health state
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Cheapest drug
Most expensive drug
ARB
Losartan (100mg): £25.94
Losartan (100mg): £25.94
Valsartan (320mg): £263.71
B
Atenolol (100mg): £13.17
Atenolol (100mg): £13.17
Acebutolol (800mg): £485.45
C
Amlodipine (10mg): £18.64
Amlodipine (10mg): 18.64
Isradipine (10mg): £431.22
D
Bendroflumethiazide (2.5mg):
£11.86
Bendroflumethiazide (2.5mg):
£11.86
Xipamide (20mg): £50.74
The cost of diuretics are also analysed in a further sensitivity analysis using the cost for: chlortalidone
(50mg(a)): £19.81; indapamide (2.5mg): £16.03.
a) 25mg was considered the optimal dose but a cost for this tablet size was not listed in the BNF.
Source: British National Formulary 60, September 2010306
I.3.8
Quality of life (utility)
In the NICE reference case, the value of health outcomes – including beneficial and harmful
impacts of treatment on mortality and morbidity – is estimated using the QALY approach.
This requires estimates of survival and quality of life associated with each health state
included in the model.
The utility values used in the model are shown in Table 92 and Table 93. An extensive
literature search was conducted during the development of the statins HTA model to identify
the best available utility estimates for the various health states625. Thus estimates for MI,
unstable angina and stroke were taken from the statins HTA. Diabetes and heart failure
estimates were taken from the Cost-Effectiveness Analysis (CEA) Registry264. For MI and
unstable angina a higher utility was applied after the initial 6 months. For diabetes, stroke and
heart failure a constant utility from onset of the condition was assumed.
410
Update 2011
Used in basecase analysis
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Table 92: Health state utility weights
Health state
Utility weight
Source
MI (first 6 months)
0.76
Statins model625
Post MI
0.88
Harvard CE Registry264
Unstable angina (first 6 months)
0.77
Statins model625
Post unstable angina
0.80
Assumption
Stroke
0.63
Statins model625
Diabetes
0.90
Harvard CE Registry264
Heart failure
0.71
Harvard CE Registry264
Death
0.00
Statins model625
Table 93: Utility weight by age
Age group
Age utility weight
Source
45–54
0.85
DH Health Survey for England 199624
55–64
0.79
DH Health Survey for England 199624
65–74
0.78
DH Health Survey for England 199624
75+
0.73
DH Health Survey for England 199624
DH = Department of Health.
As in the Statins model625, utilities were adjusted to reflect the fact that health-related quality
of life in the general population decreases with age (that is, multiply the disease utility weight
by age utility weight). Age utility weights were taken from the Department of Health, Health
Survey for England (1996)24.
Antihypertensive medication may be expected to have two opposing effects on quality of life:
improvements through the reduced incidence of CVD events (as discussed above) and
reductions through the impact of treatment-related adverse effects. The latter could potentially
be important in assessing the balance between benefits and harms, particularly for low-risk
individuals. Differences in adverse effects between the drugs could also have an influence on
their relative cost effectiveness. A Medline search was done to identify utility estimates that
could be used to reflect the latter for the included drug classes. Some studies were identified
that estimated the incidence of drug-related adverse events and quality of
life129,160,161,218,278,598,607,662. However, none of these included data in a form suitable for
estimation of utilities. Most published cost-effectiveness studies have assumed zero, or
minimal (0.01), loss of quality of life due to treatment-related side effects (Harvard CEA
Registry264). Where these have compared different antihypertensive medications, they have
generally assumed equal utility loss from adverse effects of treatment301,307. Few studies have
directly measured treatment utilities from patients. The economic analysis of the SCOPE trial
included direct assessment of utility using the EuroQoL health status measurement
instrument166. This estimated a mean change in utility of minus 0.03 for the candesartan group
and minus 0.05 for the mixed hypertensive treatment control group over a mean follow-up of
3.7 years. However, it is not possible to separate out the impact of treatment side effects, or to
attribute utility losses to individual drugs. Another cost-effectiveness study412 estimated
utilities from 148 hypertensive patients using the standard gamble technique. They found a
net loss in utility of 0.027, but did not report any difference by drug.
Given this paucity of information, we assumed no loss of utility due to adverse effects of the
drugs in the basecase model. However, we did a sensitivity analysis to investigate how large
any such effects would have to be to change the results.
411
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
I.4
Cost effectiveness
The results of cost-effectiveness analysis are usually presented as incremental costeffectiveness ratios (ICERs), which determine the additional cost of using one drug (X) per
additional QALY gained compared with no intervention or another drug (Y).
Cost of X − Cost of Y
ICERs =
QALY of X − QALY of Y
Where more than two interventions are being compared, the ICERs are calculated using the
following process:
1. The drugs are ranked in terms of cost, from the cheapest to the most expensive
(cheapest indicated by LC (lowest cost) in the results tables below).
2. If a drug is more expensive and less effective than the previous one, then it is said to
be ruled out by 'simple dominance' and is excluded from further analysis (indicated by
a dash ‘-‘ in the results tables below).
3. ICERs are then calculated for each drug compared with the next most expensive nondominated option. If the ICER for a drug is higher than that of the next most effective
strategy, then it is ruled out by 'extended dominance' (indicated by a dash ‘-‘ in the
results tables below).
4. ICERs are recalculated excluding any drugs subject to dominance or extended
dominance (these ICERs are given in the results tables below).
I.5
Sensitivity analysis
The model includes a base-case analysis supplemented with univariate and multivariate
deterministic sensitivity analyses to test the impact of uncertainty over various model
parameters and assumptions.
I.6
Results
I.6.1
Base-case results
The base-case results are presented in Table 94 for 65-year-old men and women with an
annual CVD risk of 2%, heart failure risk of 1% and diabetes risk of 1.1%. This suggests that
antihypertensive treatment is cost effective for this population and that the most cost-effective
initial drug class in this group is calcium-channel blockers (C). The ICER of C compared with
thiazide-type diuretics (D) is £1,520 to £1,960 per QALY gained, which is below the level
usually considered to be affordable in the NHS (about £20,000 to £30,000 per QALY).
Beta-blockers (B) are ruled out by simple dominance, since D, A and C are estimated to be
cheaper and more effective. This is illustrated in Figure C2, since B lies to the northwest of D,
A and C. The ACEi/ARB option (A) is also ruled out by extended dominance, since treating
some patients with D and the remainder with C would be cheaper and more effective than A;
in Figure 119, A lies to the northwest of a straight line joining points D and C. However, it
should be noted that the absolute differences between A, C and D are small.
412
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Table 94: 2011 base-case results (65-year-old, 2% risk, 1.1% diabetes risk, 1% heart
failure risk)
Men
Cost (£)
Effect (QALYs)
ICER (cost per QALY)
D
£3,910
10.22
LC
A
£4,010
10.21
-
C
£4,030
10.28
£1,960
B
£4,550
9.89
-
NI
£4,690
9.57
-
Women
Effect (QALYs)
ICER (cost per QALY)
£4,310
10.65
LC
C
£4,390
10.71
£1,520
A
£4,400
10.63
-
B
£5,050
10.29
-
NI
£5,230
9.96
-
Update 2011
Cost (£)
D
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
Figure 119: 2011 base-case results (65-year-old, 2% cardiovascular risk, 1.1%
diabetes risk, 1% heart failure risk)
Mean cost (2009 UK £ per person,
discounted)
Men
Women
£5,400
£4,800
£5,200
£4,600
£5,000
£4,400
£4,800
£1,960
£4,200
£1,580
£4,600
£4,000
£4,400
£3,800
9.40
9.60
9.80
10.00
10.20
10.40
£4,200
9.80
10.00
10.20
10.40
10.60
10.80
Mean effect (QALYs per person,
discounted)
Mean effect (QALYs per person,
discounted)
No Intervention
Calcium-channel Blockers
ACE Inhibitors/Angiotensin II receptor antagonists
Thiazide-type Diuretics
Beta-blockers
I.6.2
Results for patient subgroups
Figure 120 and Figure 121 show how cost effectiveness varies with age, sex, annual CVD
risk, annual diabetes risk and annual heart failure risk, based on a cost-effectiveness threshold
of £20,000 per QALY. The meta-analysis found that thiazide-type diuretics and CCBs have
similar effects on the incidence of MI, stroke and death (Chapter 10). However, CCBs are
associated with significantly higher rates of heart failure but lower rates of diabetes. Thus,
CCBs appear to be a more cost-effective option for people over 55 years of age at relatively
low risk of heart failure and for those at relatively high risk of diabetes. The GDG noted that
in a 55 year old the upper rates of heart failure and diabetes explored in the analysis were
extremely unlikely.
413
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
A appear to be a cost-effective alternative to C at high levels of diabetes risk and intermediate
levels of heart failure risk. This is because they are associated with lower rates of heart failure
and diabetes, but higher rates of stroke.
414
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Figure 120: Four-way sensitivity analysis: most cost-effective (represented by colour)
first-line drug for men by age, annual risk of cardiovascular disease, diabetes and heart
failure, based on a cost-effectiveness threshold of £20,000 per quality-adjusted life-year
55-year-old-man
65-year-old-man
CVD risk 1%
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.5
2
4
4
4
4
5
5
HF risk %
1.0 1.5 2.0
2
2
2
2
2
2
2
2
2
5
5
2
5
5
5
5
5
5
5
5
5
3.0
2
2
2
2
2
2
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
2
2
2
2
2
4
2
2
2
2
4
2
2
2
2
4
5
2
2
2
4
5
5
2
2
5
5
5
5
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
5
2
2
4
4
5
5
2
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
2.5
2
2
2
2
2
5
5
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.5
4
4
4
4
4
4
4
HF risk %
1.0 1.5 2.0
2
2
2
4
2
2
4
2
2
4
4
2
4
5
5
4
5
5
5
5
5
3.0
2
2
2
2
2
5
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
5
5
2
4
4
5
5
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Thiazide-type diuretic
2.5
2
2
2
2
2
5
5
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.5
4
4
4
4
4
4
4
CVD risk 1%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
5
2
4
4
4
5
5
4
4
5
5
5
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
4
4
4
4
4
4
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
4
2
2
2
4
4
4
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
ACE inhibitor (80%) or ARB (20%)
Basecase analysis
ARB = angiotensin-II receptor blocker; CVD = cardiovascular disease; HF = heart failure;
Figure 121: Four-way sensitivity analysis: most cost-effective first-line drug for
women by age, annual risk of cardiovascular disease, diabetes and heart failure, based
on a cost-effectiveness threshold of £20,000 per quality-adjusted life-year.
55-year-old-woman
65-year-old-woman
CVD risk 1%
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.5
2
4
4
4
4
5
5
HF risk %
1.0 1.5 2.0
2
2
2
2
2
2
2
2
2
5
2
2
5
5
2
5
5
5
5
5
5
3.0
2
2
2
2
2
2
2
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
2
2
2
2
2
4
2
2
2
2
4
2
2
2
2
4
5
2
2
2
4
5
5
2
2
4
5
5
5
2
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
5
5
2
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
2.5
2
2
2
2
2
2
5
Calcium-channel blocker
75-year-old-woman
CVD risk 1%
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
Thiazide-type diuretic
0.5
4
4
4
4
4
4
4
HF risk %
1.0 1.5 2.0
2
2
2
2
2
2
4
2
2
4
5
2
4
5
5
5
5
5
5
5
5
3.0
2
2
2
2
2
2
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
2
2
2
2
2
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
5
2
4
4
5
5
5
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
4
Diabetes
Risk %
0.0
0.5
1.0
1.5
2.0
2.5
3.0
2.5
2
2
2
2
2
5
5
ACE inhibitor (80%) or ARB (20%)
ARB = angiotensin-II receptor blocker; CVD = cardiovascular disease; HF = heart failure;
415
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.0
4
4
4
4
4
4
4
0.5
4
4
4
4
4
4
4
CVD risk 1%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
5
5
5
4
4
5
5
5
0.5
4
4
4
4
4
4
4
CVD risk 2%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
2
2
2
2
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
4
4
4
4
4
4
0.5
4
4
4
4
4
4
4
CVD risk 3%
HF risk %
1.0 1.5 2.0 2.5 3.0
4
4
2
2
2
4
4
2
2
2
4
4
4
2
2
4
4
4
4
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Basecase analysis
Update 2011
Calcium-channel blocker
75-year-old-man
CVD risk 1%
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
In addition, as Table 95 shows, when CVD is 0.5% (holding all other variables constant at
their base-case values, for men and women age 55 A becomes the preferred option. This
suggests that A could be cost effective in in the young/low risk people.
Table 95: Sensitivity analysis for annual CVD risk and age (1.1% diabetes risk, 1%
heart failure risk)
Annual
CVD
risk
0.5%
1.0%
Age
55
65
75
85
55
65
75
85
A dominant
£16,860
£5,370
£2,950
£2,350
£2,000
£2,240
£1,970
A dominant
£3,750*
£12,440
£3,490
£2,800
£2,090
£2,330
£1,860
2.0%
£4,000
£1,960
£1,510
£1,530
£3,200
£1,520
£1,210
£1,170
3.0%
£2,210
£1,430
£1,170
£1,190
£1,050
£800
£700
£680
5.0%
£1,180
£900
£740
£730
£260
£230
£160
£70
Update 2011
Incremental cost-effectiveness ratios of most effective (highest QALYs) option
MEN
WOMEN
All ICERs above are for C apart from * where ICER is for D
I.6.3
Younger people
The model is not designed to estimate cost effectiveness for a younger population, since most
of the evidence about treatment effects derives from studies in older people. However, we can
use the model to test the possible impact of improved performance of ACEis, ARBs and BBs
in a younger, non-black group . Taking the predicted baseline effects of a 45-year-old cohort
(at 2% annual CVD risk, 1% annual heart failure risk and 1.1% annual diabetes risk), cost
effectiveness was estimated for given percentage improvements in treatment effects for
ACEi/ARB and BB compared with the meta-analysis figures.
Diuretics appear to be the most cost-effective option for this age group in the base-case
analysis, as shown in Table 96. However, if the relative risks for ACEi/ARBs were only 1%
(men)/1.2% (women) or better than the meta-analysis estimates, then they would be cost
effective (cost per QALY less than £20,000). Beta-blockers continued to be dominated even
at higher percentage improvements, assuming an equal percentage improvement of
ACEi/ARBs and BBs for the younger population. This analysis does lend some support to the
hypothesis that ACEi/ARBs may be cost effective in younger non-black patients.
Table 96: Sensitivity analysis for increased effectiveness of A/B for younger patients (45year-old, 2% CVD risk, 1.1% diabetes risk, 1% heart failure risk)
Incremental cost-effectiveness ratios (cost per QALY)
MEN
WOMEN
D
B
A
C
*
D
B
A
C
*
0%
LC
-
-
-
3
£610
-
LC
-
2
1%
-
-
LC
-
1
£5,730
-
LC
-
2
2%
-
-
LC
-
1
-
-
LC
-
1
3%
-
-
LC
-
1
-
-
LC
-
1
4%
-
-
LC
-
1
-
-
LC
-
1
10%
-
-
LC
-
1
-
-
LC
-
1
12%
-
-
LC
-
1
-
-
LC
-
1
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A dominates all
2 = B and C dominated: D versus A
3 = D dominates all
416
Update 2011
% improvement
in effects of A/B
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
I.6.4
Other sensitivity analyses
A range of sensitivity analyses were conducted to assess the impact of different input
parameters on the base-case results. In these analyses we held all other parameters fixed at
their base-case values. The results are interpreted using a cost-effectiveness threshold of
£20,000 per QALY. Table 97 summarises the results for those parameters that led to a change
of conclusion from the base case. These results are discussed further in the sections that
follow this table.
Table 97: List of sensitivity analyses results that altered base-case conclusions
Parameters changed
Most cost-effective option
Base case men age 65 (2% CVD risk, 1.1% diabetes risk, 1% heart failure
risk)
C
Treatment effects
D
Lower limits for effects of ACEi vs C
A
Lower limits for effects of ARB vs ACEi
A
Alternative treatment effect data scenario 1: lower limits for effects of B vs D
B
Alternative treatment effect data scenario 2: lower limits for effects of ACEi
vs D
A
Event risks
RR of CVD events with diabetes of 4, when risk of diabetes is 4%
A
RR of CVD events with heart failure >11 (base case heart failure risk) and any
RR >1 when when heart failure risk is 2% and 4%.
D
Quality of life
Reduction in quality of life from drug side effects 3.6% or more
NI
Reduction in quality of life of 0.27% or more due to side effects of C
D
Costs
Cost of CCBs more than £94 per annum
D
Highest cost drug in each class used
D
I.6.4.1
Uncertainty over treatment effects
The results are sensitive to uncertainty over the magnitude of treatment effects estimated from
the meta-analyses (Table 98, Table 99 and Table 100).
The conclusion that CCBs were the most cost-effective option was robust to variations in the
treatment effects, except in the following scenarios:
 Diuretics dominate all other options when the effects of CCBs compared with diuretics are
increased to their upper 95% confidence limits.
 ACEi/ARB are the most cost-effective option in three tested scenarios:
o Lower limits for effects of ACEi versus C (A dominates all interventions).
o Lower limits for effects of ACEi versus D (£620 per QALY for A versus D).
o Lower limits for effects of ARB vs ACEi (£840 per QALY for A versus D, C versus A
> £20,000 per QALY gained)
 Beta-blockers are the most cost-effective option if we take the lower limits for the effects
of B versus D (£2,010 per QALY for B versus D).
These extreme results may be relatively unlikely, however, since the relative risks for all
outcomes would all have to be simultaneously at their lower 95% limits.
417
Update 2011
Upper limits for effects of C vs D
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Table 98: Sensitivity analysis for efficacy of treatment (65-year-old men with 2% CVD
risk, 1.1% diabetes risk, 1% heart failure risk)
Incremental cost-effectiveness ratios (cost per QALY)
Comparison
Lower limit of treatment effect
Upper limit of treatment effect
D
LC
B
-
A
-
C
£970
*
1
D
LC
B
-
A
-
C
£4,470
*
1
C vs D
-
-
LC
£360
2
LC
-
-
-
3
B vs C
LC
-
-
£1,960
1
LC
-
-
£1,960
1
ACEi vs C
-
-
LC
-
4
LC
-
-
£1,960
1
ARB vs
ACEi
5
LC
-
-
LC
-
£840
£55,420
Update 2011
D vs NI
1
£1,960
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
2 = Band D dominated: C versus A
3 = D dominates all
4 = A dominates all
5 = B dominated: A versus D, C versus A
Table 99 shows how results do not change if the treatment effects for BB are taken from the
mean relative risks in comparison with diuretics (rather than compared with CCB as in the
base-case model). BBs remain dominated and CCBs are the most cost-effective option in this
case. If the lower limits of the confidence intervals for BB compared with diuretics are used,
BB appear to be the most cost-effective option with an estimated ICER of £2,010.
Incremental cost-effectiveness ratios (cost per QALY)
Comparison
B vs D
Lower limit of treatment effect
D
L
C
B
A C
£2,010
-
£1,960
Basecase treatment effect
* D
L
1
C
B
A
-
-
Upper limit of treatment
effect
C
D
£1,960
2
B
A
-
-
LC
C
£1,960
*
2
Update 2011
Table 99:
Sensitivity analysis for treatment effects (65-year-old men with 2% CVD
risk, 1.1% diabetes risk, 1% heart failure risk) (Alternative treatment effect data
scenario 1: BB versus DD)
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A dominated: C vs D, B versus C
2 = A and B dominated by C: C versus D
Table 100 shows how the results also do not change if the treatment effects of ACEi are based
on their mean relative risks compared with diuretics, rather than with CCBs as in the basecase model. However, the ACEi/ARB combination appears to be the most cost-effective
option if the lower confidence intervals for the effects of ACEi versus diuretics are used.
Incremental cost-effectiveness ratios (cost per QALY)
Comparison
ACEi vs D
Lower limit of treatment
effect
Base-case treatment effect
D
B
A
C
* D
B
A
C
L
C
-
£620
-
1 LC
-
-
1,760
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = B and C dominated: A versus D
2 = A and B dominated: C versus D
418
Upper limit of treatment effect
2
D
B
A
C
*
LC
-
-
£1,760
2
Update 2011
Table 100: Sensitivity analysis for treatment effects (65-year-old men with 2% CVD
risk, 1.1% diabetes risk, 1% heart failure risk) (Alternative treatment effect data
scenario 2: ACEi versus DD)
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
I.6.4.2
Use of ARBs
The percentage of ARBs used in conjunction with ACEi in the base-case model was assumed
to be 20%. The model is not sensitive to assumptions about the number of patients who
cannot tolerate ACEis and switch to ARBs. Varying this percentage up to 100% did not
impact conclusions; CCBs remained cost effective (Table 101).
Table 101:
Sensitivity analysis, percentage of ARBs used in conjunction with ACEi
(65-year-old men, 2% CVD risk, 1.1% diabetes risk, 1% heart failure
risk)
WOMEN
D
B
A
C
0%
LC
-
-
£1,960
LC
-
-
£1,520
10%
LC
-
-
£1,960
LC
-
-
£1,520
15%
LC
-
-
£1,960
LC
-
-
£1,520
20%
LC
-
-
£1,960
LC
-
-
£1,520
25%
LC
-
-
£1,960
LC
-
-
£1,520
50%
LC
-
-
£1,960
LC
-
-
£1,520
60%
LC
-
-
£1,960
LC
-
-
£1,520
70%
LC
-
-
£1,960
LC
-
-
£1,520
80%
LC
-
-
£1,960
LC
-
-
£1,520
90%
LC
-
-
£1,960
LC
-
-
£1,520
100%
LC
-
-
£1,960
LC
-
-
£1,520
Update 2011
Incremental cost-effectiveness ratios (cost per QALY)
% of
MEN
ARBs
D
B
A
C
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
I.6.4.3
Event risk assumptions
Risk of CVD events for people with diabetes
In the base-case analysis it is assumed that the relative risk of CVD event for people with
diabetes is 2 (that is the risk of CVD events is doubled). At all levels of diabetes risk, CCBs
remained the most cost effective option in most scenarios regarding the relative risks of CVD
events; even if the relative risk of CVD events with diabetes was set to 1 (that is, no increase
in risk of CVD events in people with diabetes). It is more cost-effective to treat with
ACEi/ARB at a high level of diabetes risk (4%) if the relative risk of CVD events is also high
(RR = 4).
Table 102:
Sensitivity analysis for relative risk of CVD events and diabetes (65-yearold men 2% CVD risk, 1.1% diabetes risk)
Incremental cost-effectiveness ratios (cost per QALY)
Annual risk of diabetes =
1.1%
Annual risk of diabetes = 2%
Annual risk of diabetes = 4%
D
B
A
C
* D
B A
C
* D
B A
C
*
1
LC
-
-
£3,760
1 -
-
LC
£1,490
2 -
-
LC
£3,360
2
2
LC
-
-
£1,960
1 -
-
LC
£1,690
2 -
-
LC
£8,620
2
4
LC
-
-
£1,640
1 LC
-
£310
£2,510
3 -
-
LC
-
4
LC = lowest cost; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
2 = D and B dominated: C versus A
3 = B dominated: A versus D, C versus A
4 = D, B and C dominated by A
419
Update 2011
RR
**
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
** Relative risk of CVD events following heart failure compared with CVD event risks for people who have not had a CVD
event
Risk of CVD events for people with heart failure
The results are sensitive to the relative risk of CVD events for people with heart failure (Table
103). For a given level of heart failure risk, the cost effectiveness of CCBs worsens as the
relative risk of CVD events for people with heart failure increases. This may be explained by
the fact that D does better in preventing heart failure than CCBs. At 1% annual risk of heart
failure (as in the base-case analysis, CCBs are no longer cost-effective compared with
diuretics only if the risks of CVD events with heart failure are more than 11 times higher than
in the base case.
Table 103:
Sensitivity analysis for relative risk and incidence of CVD events following
heart failure (65-year-old men 2% CVD risk, 1.1% diabetes risk)
Annual risk of heart failure =
4%
D
B
A
C
*
1
LC
-
-
£1,960
1
LC
-
-
-
2
LC
-
-
-
2
2
LC
-
-
£2,960
1
LC
-
-
-
2
LC
-
-
-
2
4
LC
-
-
£4,990
1
LC
-
-
-
2
LC
-
-
-
2
Update 2011
Incremental cost-effectiveness ratios (cost per QALY)
RR Annual risk of heart failure =
Annual risk of heart failure =
**
1%
2%
D
B
A
C
*
D
B
A
C
*
LC = lowest cost; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
2 = D dominates all other options
** Relative risk of CVD events following heart failure compared with CVD event risks for people who have not had a CVD
event
Risk of CVD events for people with stroke
As shown in Table 104, conclusions are not sensitive to the relative risk of CVD events
following a stroke; CCB remains the most cost-effective option.
Table 104:
Sensitivity analysis for relative risk of CVD events following a stroke (65year-old 2% CVD risk, 1.1% diabetes risk, 1% heart failure risk)
WOMEN
D
B
A
C
*
D
B
A
C
*
1
LC
-
-
£1,960
1
LC
-
-
£1,520
1
2
LC
-
-
£1,970
1
LC
-
-
£1,580
1
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
** Relative risk of CVD event following stroke compared with CVD event risks for people who have not had a CVD event
Update 2011
Incremental cost-effectiveness ratios (cost per QALY)
RR** MEN
Risk of non-CVD death
As shown in Table 105, conclusions are not sensitive to changes in the assumptions about the
relative risk of death from non-CVD in the hypertensive cohort compared with the general
population. Hypertensive treatment remains highly cost-effective, and CCBs remain the
preferred option (holding all other variables at their base-case values).
Table 105:
Incremental cost-effectiveness ratios (cost per QALY)
RR**
MEN
WOMEN
420
Update 2011
Sensitivity analysis for relative risk of non-CVD death (65-year-old 2%
CVD risk, 1.1% diabetes risk, 1% heart failure risk)
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Incremental cost-effectiveness ratios (cost per QALY)
D
B
A
C
* D
B
A
C
*
1
LC
-
-
£1,960
1
LC
-
-
£1,520
1
2
LC
-
-
£1,500
1
LC
-
-
£1,310
1
4
LC
-
-
£1,250
1
LC
-
-
£1,160
1
8
LC
-
-
£1,030
2
LC
-
-
£1,000
1
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 =A and B dominated: C versus D
2= A and B dominated: C versus D and D versus NI
** Relative risk of non-CVD death for population in model (people with hypertension) compared to general population
I.6.4.4
Quality of life
Quality of life due to drug side effects
The base-case model assumes there is no loss in quality of life due to hypertensive treatment
side effects. If the loss of quality of life due to the side effects of hypertensive treatment is
assumed to be 3.6% or greater, then treatment may not be cost effective. This assumes equal
quality of life loss for all drugs, which is unlikely given that we know that there are differing
rates of adverse events and withdrawals.
Table 106:
Sensitivity analysis for quality of life loss from hypertensive treatment (65year-old, 2% CVD risk, 1.1% diabetes risk, 1% heart failure risk)
Incremental cost-effectiveness ratios (cost per QALY)
MEN
WOMEN
D
B
A
C
*
D
B
A
C
*
0%
LC
-
-
£1,960
1
LC
-
-
£1,520
1
1%
LC
-
-
£1,960
1
LC
-
-
£1,520
1
2%
LC
-
-
£1,960
1
LC
-
-
£1,520
1
3%
LC
-
-
£1,960
1
LC
-
-
£1,520
1
4%
LC
-
-
£1,960
2
LC
-
-
£1,520
2
5%
LC
-
-
-
3
LC
-
-
£1,520
2
Update 2011
QoL
reduction
LC = lowest cost; QoL = quality of life; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
2 = A and B dominated by NI: NI versus C (NI ICER not shown – NI cost effective)
3 = A, B and C dominated by NI: NI versus D (NI ICER not shown – NI cost effective)
Small differences in adverse effects of the different drugs may change their relative cost
effectiveness. Holding all other parameters constant at their base-case values, CCBs remain
the most cost-effective option provided that their impact on quality of life due to adverse
effects does not exceed about 0.27% (Table 107). For comparison, the quality of life impact
of chronic lower-extremity oedema has been estimated at 10% (Harvard CEA registry). Thus,
if an individual experiences even minor or infrequent side effects with CCBs, then alternative
antihypertensive treatment may be more cost effective.
Update 2011
Table 107:
Sensitivity analysis for quality of life with CCBs and ACEi/ARBs (65year-old men, 2% CVD risk, 1.1% diabetes risk, 1% heart failure risk)
Incremental cost-effectiveness ratios (cost per QALY)
Reduction
with C
0.1%
0% reduction with
A
0.2% reduction with
A
D
A C
* D
A
L
C
-
1 L
C
-
£2,940
0.4% reduction with
A
C
* D
A
1 L
C
-
£2,940
421
C
£2,940
0.6% reduction with
A
D
1 L
C
A
C
-
*
1
£2,940
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Incremental cost-effectiveness ratios (cost per QALY)
0.2%
L
C
-
1 L
C
-
0.4%
L
C
-
-
2 L
C
-
0.8%
L
C
-
-
2 L
C
1.0%
L
C
-
-
2.0%
L
C
-
-
£5,890
1 L
C
-
-
2 L
C
-
-
-
2 L
C
2 L
C
-
-
2 L
C
-
-
£5,890
1 L
C
-
-
2 L
C
-
-
2
-
-
2 L
C
-
-
2
2 L
C
-
-
2 L
C
-
-
2
2 L
C
-
-
2 L
C
-
-
2
£5,890
1
£5,890
LC = lowest cost; ’-‘ = option ruled out by simple or extended dominance
* 1 = B and A dominated: C versus D
2 = D dominated all
Quality of life due to events
Table 108 interpretation: The results are not sensitive to changes in the assumed quality of life
change due to CVD events or the onset of diabetes. C remained the most cost-effective option
under all scenarios tested.
Table 108:
Sensitivity analysis for quality of life loss from CVD events and diabetes
(65-year-old men, 2% CVD risk, 1.1% diabetes risk, 1% heart failure
risk)
Lower limit
Upper limit
Quality of life loss
D
B
A
C
* D
B
A
C
*
Unstable angina (0.7-0.9)
LC
-
-
£1,960
1 LC
-
-
£1,950
1
MI (0.7-0.9)
LC
-
-
£1,960
1
LC
-
-
£1,950
1
Diabetes (0.8-1)
LC
-
-
£1,930
1
LC
-
-
£1,990
1
Stroke (0.5-0.7)
LC
-
-
£1,930
1
LC
-
-
£1,960
1
Heart failure (0.6-0.8)
LC
-
-
£2,010
1
LC
-
-
£1,920
1
Update 2011
Incremental cost-effectiveness ratios (cost per QALY)
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
I.6.4.5
Costs
Drug costs
In the base-case model, CCBs were assumed to cost £18.64 per patient per annum (based on
the BNF 60, September 2010, price of amlodipine). If this is increased to £94 or more, then
CCBs were no longer cost effective compared with diuretics.
As shown in table Table 109, the model is sensitive to assumptions about the cost of drugs.
CCBs remained the most cost-effective option when the cheapest drugs are used. When the
most expensive drugs are used, the ICERs increase to a level above what is usually considered
affordable by the NHS, between £20,000–£30,000 per QALY, making D the optimal choice.
However, this is an unlikely scenario.
Table 109:
Incremental cost-effectiveness ratios (cost per QALY)
Cost of
drugs
MEN
D
WOMEN
B
A
C
* D
422
B
A
C
*
Update 2011
Sensitivity analysis for cost of drugs (65-year-old, 2% CVD risk, 1.1%
diabetes risk, 1% heart failure risk) (cheapest and most expensive)
Hypertension (partial update)
Cost-effectiveness analysis – pharmacological treatment (updated 2011)
Incremental cost-effectiveness ratios (cost per QALY)
Cheapest
LC
-
-
£1,960
1 LC
-
-
£1,520
2
Most
expensive
LC
-
-
£85,160
1 LC
-
-
£95,660
1
LC = lowest cost option; ’-‘ = option ruled out by simple or extended dominance
* 1 = A and B dominated: C versus D
2 = B dominated: C versus A
As shown in Table 110, the conclusion that CCBs are the